Skip to main content

Full text of "New Energy Times Reprint of the Bhabha Atomic Research Centre 1500 Report"

See other formats

Preface to the New Energy Times Reprint of the BARC-1500 Report 

It is appropriate that an international conference on "cold fusion," now called Low 
Energy Nuclear Reactions (LENR) is being held next month at Salt Lake City 
Utah, to mark the twentieth anniversary of the historic Fleischmann-Pons 

University of Utah announcement which was first made there. 

It is obvious by now that experiments on this field hay^ been repeated by several 
groups in the world and there is nothing fundamentally wronq with the 

/• * 

If the observed phenomena do not fit within our text-book understanding of 
nuclear phenomena, that is a problem for science to solve. Nature demonstrates 

many phenomena which we don't yet understand. This does not mean we should 
not explore further. 

The need to satisfy peer reviews,- (in fact there could be no perfect peers at any 
'me in any subject,) should not come in the way of continued exploration. We are 
glad that an archive is being created to mark this occasion. We congratulate 
Steven B. Krivit and New Energy Times for this initiative. 

- Dr. P. K. Iyengar, Chairman (retired), Atomic Energy Commission India 
February 23, 2009 

" M - Snn i v ¥# Associate Director (retired), Physics Group, Bhabha Atomic 
Research Centre (BARC), India 

February 23, 2009 

B. A.R.C.- 1500 


( as pet IS ; 9400 - 1980 ) 



RT A: 
A 1 

A 2 



.* ■ 

Cold Fusion Experiments Using a Commercial Pd— Ni Electrolyser 
M S. Krishnan, S.K. Malhotra, D G. Gaonkar, M. Sriuivasan, 

S.K. Sikka, A. Shyarn, V. Chitra, T.S. Iyengar and PET. Iyengar 

Preliminary Results of Cold Fusion Studies JO ssr ngHC Five M 
Current Electrolytic Cell 
M.G. Nayar, S.K. Mitra, P. Raghunathan, M.S. Krishnan, S.K. Malhotra, 
D.G. Gaonkar, S.K. Sikka, A. Shyarn and V. Chitra 

A 3 Observation of Cold Fusion m a Ti-SS:J| ; fectrolytic Cell 

M.S. Krishnan, S.K. Malhotra. D G. Gaonkar, M.G. Nayar, A. Shyarn and 


S.K. Sikka 

A 4 Multiplicity Distribution of Neutron Emission in Cold Fusion Experiments 
A. Shyam, M Srinivasan, Degwekar and L.V. Kulkarni 

A 5 Search for Etectrocl^ihicaily Catalysed Fusion, of Deuteroos in Metal 

r.P. Radhakrishiian,: R. Sundaresan, J. Arunachalam, V. Sitarama Raju, 

R Kalyanaramap, S Gangadharan and P.K. Iyengar 

A 6 Tritium. Generation during Electrolysis Experiment 

T P. Radhaknshnan, R. Sundaresan, S. Gangadharan, B.K Sen, 


A? Burst Neutron Emission and Tritium Generation from Palladium Cathode 
Electrolyticaily Loaded with Deuterium 

G: Venkateswaran, P.N. Moorthy, K.S. Venkateswarlu, S.N. Guha, 8 
Yuvaraju, T. Datta, T.S. Iyengar, M.S. Panajkar, K.A, Ran and Kamal 

A 8 Verification Studies in El e cfcr ochemically Induced Fusion of Deuterons in 
Palladium Cathodes 

H.. Bose, L.H Pr&hhu, S. Sankarnarayanan., R.S. Shetiva, N. Veeraraghavan, 
P.V. Joshi, T.S. Mur thy, B.K. Sen and K G-B Sharma 

A 9 Tritium Analysis of Samples Obtained from Various Electrolysis 
Experiments at BARC 

T.S. Murthy, T.S. Iyengar, B.K. Sen and T.B. Joseph 

A 10 Material Balance of Tritium in the Electrolysis of Heavy Water 
M.S. Krishnan, S.K. Malhotra and S.K. Sadhukhan 

A 11 Technique for Concentration of Helium in Electrolytic Gases for Cold Fusion 
K.Aimaji Rao 

CONTENTS (Contd.) 


B 1 Search, for Nuclear Fusion in Gas Phase Denteriding of Ijtipium Metal. 
P.ftaj, P. Suryanarayana, A. Sathyamoorthy and T. Datfca 

B 2 Deuteration of Machined Titanium Targets for Cold Fusion." Experiments 
V.K. Shrikande and K.C. Mittal 

B 3 Autoradiography of Deuterated Ti and Pd TuMrfor Spatially 'Resolved 
Detection of T rithun Produced by Cold Fusion \ ' * ■ * 
a.K. Rout, M. Srinivasan and A. S.hyam 

B 4 Evidence for Production of Tritin^yrie Cold Fusion Reactions in 
Deuterium Gas Loaded Palladium 

M.S. Krishnan, S.K. Malhotra, D.G. Gaonkar. V.D- Nagvenkar and 
H.K. Sadhukhan f 


C 1 Materials Issues in the .So — Called *Cold Fusion* Experiments 
R Chidambaram and V.C. Sahni 

C 2 Remarks on Cold Fusion 

B.A. Dasannacharya and K.R. Rao 

C 3 The Role of Combined Electron— Deuteron Screening in D— D Fusion in 
Metals 'V 

S. N. Vaidya and Y.S. Mayya 

C 4 ^ A Theory of Cold Nuclear Fusion in Deuterium Loaded Palladium 
Swapan K. Ghosh, H.K. Sadhukhan and Ashish K. Dhara 

C.5 Fracture Phenomena in Crystalline Solids: A Brief Review in the Context 
of Cold Fusion 

T. O. Kaushik, M. Srinivasan and A. Shyam 


Energy production in the Universe is mostly based on nuclear reactions especially 
fusion reactions of light dement nuclei. Energy production in the Sun on the basis of fusion 
of hydrogen, its isotopes and elements upto carbon have been well theorized by now. It is 
natural to expect that there will be. a large variety of nuclear reactions which will lead to 
the .production of nuclear energy. In fact there is a whole gamut of fusion reactions in. 
astrophysics which suggest various combinations of nuclear interactions ami modes of decay 
for energy productioniThe collapse of binary stars and the transformation oi neutron stars 
into black holes are the ultimate phases of stellar evolution and production of fusion energy 
therefrom. Even before the discovery of the neutron, scientists had predicted and even tried 
to prove that nuclear energy could be generated through fusion of hydrogen nuclei 
i protons). It was however only after detailed .accelerator based research m. nuclear physics, 
that the cross sections and Q values for these reactions becarnfe^available. This enabled 
many conjectures to be made. Some of the candidate rlactions considered for the 
generation of fusion energy are; (p+p), (p+d), (d+d), (d4'l)-,.etc . 

The familiarity of scientists with accelerator based nuclear reactions, however led 
them to believe that fusion reactions can take place ’only on the basis of overcoming the 
potential barrier caused by the electrostatic interaction This demands that the particles 
have considerable relative velocity and from%|elnalogy of what is happening in the Sun, 
thermonuclear fusion was considered as the molt appropriate technique for releasing fusion 
energy on a large scale. We are all aw^mof the many experimental attempts that have 
&een .made over the last- four decades to obtain conditions appropriate for thermonuclear 
fusion. The principle of confinement bfithese hot plasmas by means of complex magnetic 
fields in special configurations was invented in the early fifties. Of these the Tokamak has 
been a major success and has .almost reached the stage of breakeven in energy production. 
However, the large size ajjd the expensive equipment, needed to attain even this breakeven 
stage have raised doub%,;gbpiit its commercial viability. The technique of generating 
temperatures of 1.00 mjlliSu/iegrees using the principle of inertial confinement was first 
demonstrated in a thermonuclear bomb. The same principle has been borrowed and 
adapted in m.aking: ; fusion reactions possible in small pellets using lasers, electron beams 
and heavy ion beams However even, this approach of releasing nuclear energy through a 
series of fusion micro— explosions has not lived upto its early expectations as the power and 
energy of the driver beams for obtaining the requisite pellet energy gain became 
uncomfortably "high. Small and more elegant methods are therefore being attempted. 
Techniques such as Z— pinches, combined, magnetic and inertial, confinement schemes etc 
are under experimentation. As an interim measure it has meanwhile been suggested that 
fusion devices may be employed as a source of neutrons for producing fissile fuel for use in 
fission reactors Thus in quest of establishing the best method of producing fusion energy 
there has been considerable innovations and cross fertilization of new ideas during the last 
couple of decades. However new ideas are always welcome and must be tried out. 

{"'old fusion winch we are discussing here is one such innovation which on the face of 
it looks so simple that it seems too good to be true. It has generated considerable 
speculation on the processes which cause fusion rn (he solid state at room temperatures. 
The basic problem is essentially to bring together ions of hydrogen isotopes at distances of 
a few Pwmia so that fusion takes place. It is worth .recalling here previous attempts to 
bring together hydrogen nuclei to distances at which the spontaneous fusioning rate would 
increase considerably. The most, effective method has been the replacement of the orbital 
electron of a molecular hydrogen ion by a p meson or muon as it is called. Because of its 
heavier mass, the muon is able to squeeze the nuclei into a more compact molecule and 
cause a fusion reaction. Besides, the muon m found to have m additional advantage, 
because of its longer life time { 2 as), freed after a fusion reaction, it is able to catalyze 


more fusions. Almost 200 catalyzed fusions per muon have been experimentally observed in 
(d— t) mixtures to date. It is some of the same scientists who were concerned with the 
physics of muon catalyzed fusion who have now reported cold, fusion m a lattice of 

Prom the point of view of understanding the physics behind cold fusion, one needs 
to discuss the lattice structure and its rearrangement when hydrogen is absorbed in 
palladium, titanium or other alloys used for the storage of hydrogen .'"it is the fact that 
enormous quantities of hydrogen can be stored in these materials at densities comparable 
to or higher than that of liquid hydrogen that first gave a due that perhaps the 
intemuclear distances can be brought down in such lattices. Many attempts have been 
made by theorists to evaluate the fusioning rate in such lattices. Some of these are based on 
an extension of the well known theory for muon catalyzed fusion* wherein both the 
intemuclear distances and the height of the potential barrier are varied. Both. will, have the 
effect of increasing the fusion rate from IQ" 64 per second^/per ion pair in heavy wafer to 
something of the order of 10~ 23 per second per ion pa|j v which is required in order to explain 
the exper imentally observed neutron production rate In cold fusion experiments. Whether 
it is possible to have such drastic changes in ..the'' .'-fusion probability, which is essentially 
dictated by quantum mechanics considerations'll ad-matter of intense debate and discussion. 

I would like to invite your attention to a novel application of the principles of 
quantum mechanics to such a problem... Several years ago Rand McNally had speculated on 
the feasibility of the occurrence of nuclear reactions at room temperature and was perhaps 
the first person to coin the phrase “cold fusion” as early as in. 1983. To quote his own words 
The problem of neutron transfer media is no longer an elementary binary collision 

process involving Coulomb barriers and brief collision times but rather one in which the 
nuclei are continuously mqgpeh other's proximity. Since 13 ^Xe has a slow neutron capture 
radius approaching that bC the inter atomic distance the nuclear barrier would perhaps be 
grossly reduced. Thuh ; .it i% Temotely possible that some combination of natural processes 
may permit barrier p§i$f6fation to occur much more readily and a nuclear reaction to 
ensue". He also purposed the term "de Broglie interaction length" to emphasize the fact 
that the de Broglie wave length of particles with small kinetic energy is very long. The 
importance of the de Broglie interaction length can be seen in the extraordinarily high 
cross section for absorption of slow neutrons by certain nuclei. It is therefore of interest to 
know what the de Broglie interaction length of a deuteron with very small energy is in a 
palladium lattice. Further, what, happens to the charge distribution of such a deuteron 
extended m space and the effects of its polarization are too speculative. If the charge 
distribution has dimensions of the order of a de Broglie interaction length, then the 
potential barrier due to Coulomb interaction may perhaps become much smaller. If so then 
fusion should be much more probable at very low temperatures. 

From the experimental point of view, the proof of cold fusion must come from a 
demonstration of the production of neutrons, He* 3 , He 4 , tritium, gamma. rays and other end 
products of nuclear transmutation reactions. Unfortunately experiments performed so far 
have used very small electrodes and small cells, and there have not been sufficiently large 
sized experiments which car. give unqualified, proof of the number of neutrons or 
radioactivity produced from this process. Our attempts in different groups at Trombay 
have however all shown reliable data on neutron and tritium production. These are 
described in the various papers included m this compilation. It is interesting that fusion 
reactions also take place when deuterium ions axe introduced into a metallic lattice by 
simply absorbing deuterium gas into titanium or palladium. The group at Frascati in Italy 
first succeeded in producing neutrons by this method. 


^ ,! t expectation that cold fusion will become an energy source tomorrow or 

the day after. After all even in 1939 when neutron induced fission m uranium was 
discovered, it took several years to find out how to set up a fission chain reaction and 
release fission energy m a large scale. Without detailed measurements of the .number of 
neutrons produced m fission by the Columbia University group and the invention by Fermi, 
oi the heterogeneous neutron multiplying system, the nuclear reactor would not have 
become a reality. Similarly one has to explore and understand the basic mechanisms of 
tuston m a lattice and determine how this could be used either to produr e energy directly 
or to produce neutrons and tritium in a sustained manner. It is too early to predict the 
tune frame in which this will happen but for those of us familiar Ifth the historical 
evolution of nuclear technology, one can foresee how it can change our perspective 
drastically it is therefore necessary for us to involve ourselves deeply into understanding 
tne mysteries oi this new phenomenon. The source of such energy, namely deuterium, is 
ordinary water which is available m plenty and the technology fp separate and concentrate 
deuterium Irom water is by now' well established in our country. 

the fr 

tnat in a couple oi decades from now fusion will become possible and that will 

ensure energy production for man’s needs aalpS^ % there is sea water on this earth** It .is 
therefore a historical occasion fox us to renew oilr efforts in research and development in an 
area so vital for human prosperity. I hope- this report will stimulate the interest of 
scientists and engineers from various diJgiplines in this centre to channelise their efforts in 
a way that we lead in this emerging technology. 

i Ins Preface is an edited version of the Inaugural talk delivered by Dr. P.K. Iyengar 
Jirector, BARG at .the one day meeting on 'Cold Fusion* held at Trombay on 18 th May 



.This report Is a compilation of the work carried out at BARG, Trombay during the 
first six months of the “cold fusion era” namely April to September 1989.1 Over fifty 
scientists and engineers besides a large number of technicians from more than, ten Divisions 
of this Centre have been associated with these studies.. This report comprises oftthree parts. 
Part A _ covers cold fusion investigations based on the electrolytic appjjfefo Part B 
summarises the work based on D 2 loading in the gas phase and Part 0 covers the 
theoretical papers. Since most of the theoretical papers have either been published or are in 
the process of being published in scientific journals they are not commented upon in this 

iecirolysis Experiments 

»v i 

Several groups having expertise in various areas such as hydridmg of metals 
electrochemistry, isotope exchange processes in the epppeAration of heavy water and. 
neutron and tritium measurements, devised and set i|p a variety of electrolytic calls. In the 
initial experiments the emphasis was on detection;' 1 '# ^neutrons and tritium rather than 
excess heat. In a centre such as BARC the. equipment and expertise for neutron 
measurements was readily available. levels m the D*0 electrolyte after 

decades. Well known liquid scintillation counting' techniques applicable for tow "energy beta 
emitters were used, taking adequate precautions as described in detail in Paper A9. Prior 
to the commencement of electrolysis* samples of the initial electrolyte also were analysed 
tritium content. 

. Tile flrst bursts of cold fusion neutrons were detected at Trombay on 21st April 
1989 m two separate electrolysis experiments. At the Heavy Water Division a readily 
avadabie commercial (Mfttpri-Roy) electrolytic cell with Pd-Ag alloy tubes as cathodes 
and Ni as anode, originally meant for the generation of H 2 gas was adapted for the 
electrolysis of 5M NaOD in D 2 0 (see Paper Al). Both a bank of BP 3 detectors embedded 
in paraffin and a proton recoil fast neutron detector (NE 102A) were employed to look for 
possible neutron emission. An initial burst of neutrons was detected when the cell current 
attained 60 amps. Later the current increased to 100 amps and the cell became overheated, 
resulting in the built-in trip circuit automatically switching off the power. This was 
followed by a big burst of neutrons approximately two orders of magnitude larger than 
background levels during a five minute interval. Fig .2 of Paper Al shows plots of the 
neutron counts data from the two detectors. Prom the efficiency of the neutron detection 
system measured using a standard Pu-Be source, it was surmised that ~ 4x1 0 7 neutrons 
were emitted in all by this cell. 

On the same day another cell set up by the Analytical Chemistry Division (ACD) 
produced neutron bursts (see Pig. 7 of paper AS). Table I presents a summary of the 
six successful electrolytic experiments conducted at Trombay wherein both neutron and 
tritium production has been confirmed. These cells have employed a variety of 
cathode-anode configurations. While some groups selected NaOD as electrolyte, others 
used LiOD. One of the successful cells has deployed, titanium as cathode mate rial . 
However many cells did not give positive results and these are not included in Table I. 

The mast 
surprisingly low 
the six successful, 
ratio, while one 

important result to emerge out of the Trombay measurements is the 
overall neutron, to tritium yield ratio. As evident from Table I five out of 
1 Trombay cells have given a value in. the region of IQ’® to HT* for this 
(WGD) has given a comparatively large value of ~1(H for this ratio. 


However it must be pointed out that in this cell the absolute amount of tritium generated 
was very small On the whole however the Trombay results dearly demonstrate that cold 
fusion is essentially "aneutronic" in nature. 

It is worthwhile carefully scrutinising Table I to see what additional information can 
be extracted from these results. The current density m all the cells (except the last) was in 
the region of 170 to 800 mA/cm 2 and the total charge passed per cm 2 of cathode area prior 
to the detection of the first neutron burst (the “switching on charge" as it may be called) 
was m the range of 0.6 to 3.2 amp — hrs/cm-. In the last cell however wherein the charging 
current density was only 60mA/cm 2 the switching on charge was an order of magnitude 
.higher (34 amp— hrs/cm 2 ) . Significantly however one ampere— hour or 3600 coulombs 
approximately corresponds to the charge carried by the number ofdeuterons required to 
load a few grams of Pd (associated with each cm 2 of cathode area.} to a D/Pd ratio of “0,6. 
Thus the fact that the switching on charge more or liss ■.-cdrrespon.ds to the defined 
deuterium loading requirement of the cathode is a noteworthy experimental finding. 

Since cold fusion in electrolytic cells appea%mibe : essentially a surface phenomenon 
the integrated neutron yield per cm 2 of cathode arehhnay of interest. This quantity lies in 
the range of 1.7 to 2.9x10® neutrons/cm 2 for four of the cells. For the five module cell 
(Paper A2) it was an order of magnitude sjmall% While for the last cell it was two orders of 
magnitude higher. The specific tritium yield (pCi/cm 2 ) shows a somewhat wider scatter in 
the data. This quantity lies in the range of : 0.5 to 1.3 pCi/cm 2 for the three cells wherein 
significant amounts of excess tritium were generated in comparison to the stock D 2 0 levels. 
In the case of the titanium cathode cell although the total amount of tritium generated was 
significant (7 pCi) the specific tritium' yield is an order of magnitude smaller (0.07 
fiC i/cm 5 ), But it must be pointed out that neither the tritium lost via the gas stream nor 
that entrapped m the solid electrode were accounted for in that experiment. The specific 
tritium yield is much smaller for the last two cells wherein not much tritium was produced, 

All in all it may be concluded that the behaviour of most of the electrolytic cells is 
reasonably consistent with each other considering the rather wide disparity in their designs, 
except perhaps for the last cell (WCD) whose behaviour is analamolous in many respects. 
The reason for this discrepancy is not immediately apparent. 

Statistical Characteristics of Neutron Emission 

There has been wide speculation in cold fusion literature that the observed neutron 
emission could, be due to the phenomenon of crystal cracking or fracture. In the beginning 
it was even suspected that cosmic ray produced muons could be catalyzing fusion reactions 
in the deuterated Pd or Ti cathode If these mechanisms were the source of cold fusion 
neutrons then it may be expected that neutron emission would be in multiples or bunches 
rather than one at a time. It is with a view to throwing some light on this question that the 
multiplicity spectrum of neutron emission was measured. The number of neutron pulses 
issuing from a bank of BF3 (or He 3 ) counters (embedded in paraffin) monitoring the cold 
fusion source was totalled over 20 ms sampling intervals and stored in a personal computer. 
There were 1000 such sampling intervals during a real time of 5 minutes. Data was 
simultaneously also recorded from a paraffin encased bank of thermal neutron detectors 
placed at a distance of -1.5 m from the source, serving as a background monitor. The data 
accumulated during periods of significant neutron emission were statistically analysed to 
yield the probability distribution of neutron counts The details of these measurements are 
discussed m Paper A4, It is concluded from these studies that neutron, emission essentially 
obeys Poisson distribution i.e. neutrons are mostly produced one at a time. However it is 
also found that occasionally 4 or more and even upto 20 neutron counts are registered in a 
single 20 ms interval. The background monitor has never yielded such, high multiplicity 



Target Experiments 

Following reports of neutron emission having been detected by the Frascati group 
with pressurised D* gas loaded Ti shavinp, two variants of this experiment were carried 
out at Trombay One group (Chemistry Division) followed the Fras6a% procedure with 
-20 g of cut Ti pieces and Dj gas pressures increasing upto 50 bars. These experiments are 
described m Paper Bl. The neutron detection system comprised, of 24 He 3 counters 
arranged in a well like array and having a counting efficiency of -10%. The neutron count 
rate reached a peak value of IQ 5 / 40s as compared to initial background levels of 80 per 40s 
(see Fig. 2 of Paper Bl), The neutrpn emission phase lasted fob several hours at times. 

In a second variation of the gas loading experiments described in Papers B2 and S3 
small machined targets (discs, cones, cylinders etc) of Tf metal (mass between Q.2 to 1 g) 
were individually loaded with Dj gas by indu&t^ftety heating them to 9QQ G in Do 
atmosphere at 1 Torr pressure and then switching off the power to the induction coil. D a 
gas was absorbed by the Ti target In the course of its cooling. The quantity of D % absorbed 
could be measured from the observed pressure drop. This corresponded to a gross (D/Ti) 
ratio of hardly 0.001. However it is believed that most of the absorbed D 2 gas was 
accumulated in the near surface region: A disc shaped Ti button loaded by this "procedure 
began emitting neutrons on its own. almost 50 hours after loading, producing -,10 s * neutrons 
over a 85 minute active phase (see Ffg.l of Paper A4). The background neutron counter 
did not show any increase in counts during this time 

A new result from the 8ARC work is the observation of tritium in D 2 gas loaded Pd 
possibly for the first, time. Ihese are described ia Paper B4 The tritium produced in 
these foils was extracted:' through a novel isotopic exchange technique, by keeping the 
deuterated foils in coqtac&Avith a small quantity of ordinary water for several hours. The 
tritium content iq the water was later measured through standard liquid scintillation 
techniques. r*V. * 

Autoradiography of Deuterated Ti and Pd Targets 

An important outcome of the Trombay work is the demonstration of 
autoradiography as a simple and elegant technique of establishing the presence of tritium in 
metallic samples wherein cold fusion reactions are suspected to have taken place. 
Deuterated samples of Ti or Pd were placed over standard medical X— ray films and 
exposed overnight. The radiographs of Ti discs showed about a dozen intense spots 
randomly distributed within the disc boundary, besides a large number off smaller spots, 
especially all along the periphery forming a neat ring of dots (see Fig. 1 of Paper B3). 
Repeated measurements with the same disc target with exposure times varying from 10 to 
40 hours gave almost identical patterns of spots, indicating that the tritium containing 
regions were well entrenched in the face of the titanium lattice. Spectral analysis of the 


•rav emissions from such targets using a Si (Li) detector clearly depicts the characteristic 
5 Kev) and KjS (4.9 Kev) peaks of titanium excited by the betas from tritium decay. 

In the case of Pd samples however the autoradiographs were more uniform. 
Interestingly it was found that the tritium betas {< 18 Kev energy) were unable to excite 
the characteristic. K X— rays of the Pd. The fact that even the second film of a stack of 
X— ray films exposed to the deuterated Ti and Pd samples gave a similar image, rules out 
the possibility that the image formation could be due to pressure or chemical reduction 


effects as may be suspected. 


Investigations of cold fusion phenomena carried out at Trombay during April to 
September 1989, have positively confirmed the occurrence of (d-d) fusion reactions in both 
electrolytic and gas loaded Pd and Ti metal lattices at ambient temperafJWs. Neutron 
emission has been observed at times even when the current to the electrolytic cell was 
switched off or in case of gas loaded Ti targets when no externally induced perturbation 
such as heating, cooling, evacuation etc was effected. The main findings of the Trombay 
investigations to date may be summarised as under: ' % '•'Vn> 

(a) Tritium is the primary end product of cold fusion reactions with the neutron to tritium 

yield ratio being approximately 10~ 8 , The variation of. tritium content in electrolyte 
samples taken at frequent intervals in some cells supports -the "'contention that neutron and 
tritium production occurs at the same time. , % 

|| > ifr, 

(b) Cold fusion in electrolytic cells appears to be predominantly a surface phenomenon, 
with a specific charge requirement of a little oyetM ampere-hour /cm 2 to switch on the 
neutron and tritium producing reactions. .-The fact that this roughly corresponds to the 
charge carried by the deuterons in a sample of Pd of a few grams/cm 2 loaded to a D/Pd 
ratio of 0.6 is particularly significant. 

(c) Once initiated the cathode produces typically about 10 8 neutrons/cm 2 and almost a fiCi 
of T/cm 2 , irrespective of the detailed nature of the cell design. It is imperative that we 
investigate the nature of the quenching mechanism that thwarts the ability of the cathode 
surface to sustain nuclear reactions. 

(d) Neutron emission both from electrolytically loaded Pd and gas loaded Ti is basically 
Poisson in nature i.e the neutrons are emitted one at a time. However it is not clear 
whether the neutrons ate generated in the (d-d) fusion reaction itself or whether it is 
produced in a secondary reaction involving the energetic protons or tritons. In this context 
it would be of interest to look for the possible presence of 14 Mev neutrons in cold fusion 
experiments" 1 • V 

(e) About IQ to 25% of the neutrons produced appear to be generated in bunches of over 
100 neutrons each within a time span of less than 20 ms. Viewed m the light of the neutron 
to tritium yield ratio of 10' 8 noted above, one would be obliged to conclude that a cascade 
representing approximately 10 10 fusion reactions occurs in under 20 ms. Since this appears 
very unlikely, lattice cracking or other mechanisms wherein the neutron to tritium 
branching ratio is close to unity, could be the source of such bunched neutron events. 

(f) Autoradiography of gas waded Pi and Pd targets demonstrates in a simple and elegant 
manner the presence of tritium. The experimentally deduced tritium to deuterium isotopic 
ratio in these targets is several orders of magnitude higher than in the initial stock DgO 
and as such cannot be explained away on the basis of preferential absorption of tritium by 
the Ti or Pd as may be suspected. The existence of highly localized regions (hot spots) on 
the target surface as well as all along the periphery of the disc wherein tritium is 
concentrated, points to the important role of lattice defect— sites in the absorption process, 
at least in case of titanium. 

(g) The very high probability for the tritium branch in cold (d-d) fusion reactions would 
indicate processes of neutron transfer across the potential barrier as postulated by 

Oppenheimer over half a century ago and elaborated* on more recently by Rand McNally, If 





Nummary of Successful Electrolytic Cell Experiment. Conduct atTrornbay 

N,PD N t PD 



Paper No. 

Cathode ; 




Pd— Ag 

18 tubes 

A-i A3 2 

Pd— Ag , s . Tr-,'" Pd 

# 4k ™ 

JF % • | 

Circular RyT Hollow 

sheets \ Cylinder 







Surface Area 
(cm 5 ! 




Anode Material! 

Ni ^ 

j> Porous Ni 



5M NaQD 



5M NaOD 

Volume (ml) 

, in. 


Cell Current (A) 



Current Density f 
(mA/crn 5 ) 

Total Amp.hrs^ 

Amp hrs/cm 5 ^ 






3 2 

Integrated N -Yield 



Tritium Yield 












Pt gauze 



Pt Mesh 

Pt Me«h 

5M LiOD 

0 1M LiOD 

0.1M LiOD 

01 M LiOD 






1 to 2 

6 to .8 












3 0 



3xl0 7 


1 4x10* 


1.4xlO H 

7.2x1 O' 5 

fi ?xm» 

1 Svlfttl 


I?' . ' 



M,S. Kfishnan’j S. K Maihotra*, D.G. GaonkaryM Srini v as an 4 , S. K . Sikk&t 
A, Shyarad , V, Chit rat, T.S. Iyengar* and P.K. Iyengar* 

'Heavy Water Division, 

+ Neutron Physics Division , 

■^Health Physics Division * . . 



# %v $ 

The first reports of observation of 'Cold Fusion' durj% tie electrolysis of heavy 
water' using Pd cathodes, resulted in frantic attempts in several laboratones of the 
world to duplicate these experiments and if possible improve upon them. Electrolytic 
coid fusion investigations were initiated at Trornbay,:% the first week of April '89 as a 
collaborative effort between the Heavy Water and Neutron Physics Divisions of BARC A 
commercial (Milton Boy) diffusion type Pd-Ag cathode/ Ni anode hydrogen generator 
which was readily available was employed for this purpose, after loading NaOD as 
electrolyte m place of the original NaOHf This paper gives details of the electrolyser 
characteristics, conditions of operation and the neutron and tritium measurements. 

The Electrolyser 

S' %, 1 %, 

Ihe electrolyser employed was.-. a diffusion type ultra— pure electrolytic hydrogen 
generator made by Milton Roy Company oi Ireland A A schematic view of the electrolytic 
ce . 13 s h° w h in Fig. la. The anode is of nickel and the cathode consists of specially 
activated palladium - siyef Jiloy membrane tubes. The outer nickel body of the cell 
along with a central md^fape serve as coaxial anodes. The 16 cathode tubes are 
mounted with the heMM PTFE spacers between these anode pipes as shown in Fig. lb 
and have a total wet. surface area of ~300em 2 . The cathode tubes are sealed at the top 
and open at the. hattem into a plenum through which the deuterium (or hydrogen) gas is 
drawn. ihe outlet of the gas plenum can be closed by means of a valve and the 
deuterium pressure_ allowed to build up. A pressure gauge reading upto 4 Kg/cm 2 is 
provided to read this pressure. The important cell parameters are summarised in Table 1 

The electrolyte used was 20% NaOD in D 2 0 of >99,75% isotopic purity and was 
prepared by reaction of moisture free Na with D 2 0. The oxygen generated at the anodes 
during electrolysis escapes through a vent outlet at the top of the cell. A set of baffle 
plates are provided at the fop to prevent alkali carryover into the vent. The hydrogen/ 
deuterium ions which impinge on the cathode-s under "the influence of the applied electric 
potential, diffuse through the walls of the Pd^Ag. tubes and escape into the ga b plenum. 
The ions recombine inside the^tubes to form molecular deuterium which is found to be of 
very high purity as analysed by gas chromatography. The electrolyte level in the cell is 

maintained with r.h#* h #»!«•» r,f » mn***^I otfl ^ f ~ __ .1 . S * .* •* t- 

is equipped with pressure control, a solenoid valve, electrolyte leak detector, low water 
level signal, temperature control etc. 

The inbuilt power supply provided by the manufacturers is capable of giving a 
current oi only zfO amps. During the initial runs only this was used and operation was 
F n S? comfortable current values of 60 to 62 amps. An important feature of the 

Si 1 , Ci / S! ® n however . is lfas potentially high current carrying capacity. The electrolyser was 
. .eretore connected to an external power supply capable of giving over 100 amps, during 


A 1 

subsequent runs. Current levels of -100 amps were however found to be possible only for 
short durations as the cell was getting overheated. To overcome this, a heat exchanger 
and a peristaltic, pump were incorporated enabling circulation of the electrolyte. 

Neutron. Monitoring 

Initially there were two types of neutron detectors used to monitor the neutron yield 
from the electrolytic cell. The first was a bank of 3 BF3 counters embedded in a paraffin 
moderator block. The second was a 80 mrn dia x 80 mm high recoil type plastic 
scintillator NE1Q2A sensitive, both to fast neutrons as well as high energy gammas. The 
detectors were mounted at a distance of about 10 cm from the’ cell, The counting 
efficiency of these detectors was established using a calibrated ffhjNBe neutron source. The 
counting rates of the detector outputs were totalized for S feiMtes each and printed out 
continuously on a scroll printer. During the first run there Was no separate background 
neutron monitor. In subsequent runs however a bank of three He 3 detectors, also 
embedded in a paraffin block, was installed about^d-h m from the cell to serve as 
background monitor. Also a personal computer bechme available not only to display 
graphically the count rate variations but also to accumulate and store the counts data 
registered in 20 ms intervals, with a view to subsequently perform a neutron multiplicity 
distribution analysis as described in Ref 4 

In some of the more recent runs a 'bank of specially fabricated silver cathode GM 
tubes embedded in a paraffin slab was used as an activation detector for neutrons. This 
type of neutron detector is ideally suited to measure the yield of a burst of neutrons 
produced in a time span of about IQs or lees. The neutron yield is deduced by counting 
the 24s half life of Ag n0 activity induced in the silver cathode of the GM tubes. The 
threshold sensitivity of the system for the geometry used was determined using a 
calibrated Pu—n—Be ji«u|rdn source was ~3xlQ s neutrons. 

Measurement of Tritium Bevels in, D 2 0 Electrolyte 

Measurements of the absolute levels of tritium in the D 2 0 electrolyte were carried 
out by the Tritium Group of the Health Physics Division to whom samples were sent. 
Details of the liquid scintillation counting techniques employed along with precautions 
taken by them to minimize errors due to chemiluminiscence and other interference effects 
are discussed in Ref. 5. After the initial electrolysis runs a valve was welded to the cell 
bottom to enable periodic drawal of samples of the electrolytic solution for tritium assay. 

Recently two microprocessor controlled an— line instruments for counting tritium 
activity deploying low energy sensitive scintillation fibres have been installed one in gas 
phase analysis and other for electrolytic solution counting. The development, testing and 
calibration of these instruments was carried out by the Pollution Monitoring Section of 
BARG. These instruments employ two photomultiplier tubes each in coincidence to 
suppress noise effects. 

Electrolysis Experiments and Observations 

Run No.l (21st April 1989) 

T o start wit. 
the electrolyte , 
background data. 

h the Milton Roy cell was operated with 20% NaOIl in natural, water as 
This operation war carried out for about 48 hours for collecting 
The cell was then drained, flushed with heavy water and filled with 

20% NaOD solution in D 2 0 prior to commencement of experiments on 21st April 1989. 
The ceil was operated initially at 30 amps and later the current was slowly raised to 


A 1 

60 amps corresponding to a current density of -200 mA/cm 2 . After operation under these 
conditions for about 3 hours both the neutron counters started showing bursts of neutron 
counts, well above background values, during some of the 5 minute intervals. After a 
further couple of hours of operation, both counting channels suddenly showed two- very 
large peaks and at the time of the last peak/ the current in the electrolyzer had suddenly 
increased to -120 amps on its own and the electrolyzer immediately got tripped. Later it 
was found that the PYC insulation of the electric connections between tifeSC. power 
supply and the electrolyzer had melted and even the soldering at the joi.j|tk/jftfid melted. 
The diodes of the power supply had also burnt out causing the trip. ,, 

The neutron counts data of this run axe presented in Table II and also shown 
r coiHiiiug efficiency of the Bfg Bank an<^Pl&tic scintillator were 

1106% and 0.4% respectively during this run. The fact thajf Wh. the counters show 
identical behaviour in spite of having very different neutron.. detection characteristics is 
noteworthy The total number of neutrons generated durmgtlhe four hour duration of this 
run is estimated to have been approximately 4x10" At the end of this experiment a 
sample of the electrolyte which was analyzed for irat-iufii content indicated a level of 
1.5 pCi/ml of tritium activity in comparison to the/ initial stock heavy water value of 
U.075 nCi/ml. As discussed later, this high build up. By a factor of -20,000, is far beyond 
what can be accounted for by electrolytic alone. 

Run No.2 (12th to 16th June 1989) ' 

1 he second series of electrolysis runs with the Milton Roy cell commenced during 
the first week of June 1989. The cell was drained and flushed with heavy water many 
limes mF decontamination of tnfiuirk T resh electrolyte solution prepared using unused 
heavy water was charged and left, in the cell over the weekend. On Monday 12th June a 
sample of this electroiytfe which, was analyzed for tritium content gave a surprisingly high 
tritium level of -0 32. npi/ml. This is attributed to tritium left over in the Pd— Ag 
cathodes from the 21st April run which must have transferred back into the electrolyte by 
chemical exchange. \ 

Electrolysis commenced at 11:07 hrs. The cell was initially operated at currents of 
'60 amps (current density -200 mA/cm 2 ). During this run the BF 3 bank in paraffin was 
located close to the cell. About 1.5 rn away there was a He® bank in paraffin which 
served as background monitor. The first neutron burst of this run was recorded within 
about half m. hour i.e at 11:40 hrs. About an hour later two more 5 minute counts 
indicated high neutron, levels. No more neutron bursts were observed for the next couple 
of days although cell operation was continued until 17:45 hrs on Wednesday 14th June 
when the cell was put off. But within a couple of hours of this there was a neut ron burst 
lasting 15 minutes. Samples of electrolyte were drawn periodically throughout this 
operation and sent for tritium analysis. The tritium content of the electrolyte did not 
show any increase. Rather, it decreased from 0-32 nCi/mi to 0.12 nCi/ml on 15th June 
89 The decrease is attributed to the fact, that the electrolyte containing 0.075 nCi/ml of 
tritium was added continuously for maintaining the level of the electrolyte. This 
obviously diluted the tritium content in the electrolyte. 

During the next 48 hours the neutron monitors did not show any .increase of counts. 
On the night of Friday, 16 th June more than two days after putting off the cell current 
however, a large neutron burst was recorded corresponding to a total neutron yield of well 
over 1 . 0 ® neutrons. Fig 3 shows a plot of neutron count variation during this burst. 
Detailed time structure of this burst is presented in Table VII of Ref. 4. A sample of 
electrolyte was drawn only on 23rd June to ensure that maximum amount of tritium on 
r d cathodes would by then have exchanged with the bulk DgO. This indicated a tritium 


A 1 

level of 121 nCi/ml The week long, experiment was terminated at this point but the 
electrolyser was left as such with the electrolyte m the cell and deuterium m gas plenum 
at a pressure of 1 kg/cm 1 2 3 4 5 above atmospheric pressure. After a lapse of about a month 
when the electrolyte was removed and analyzed, the tritium level was found to have 
further increased to 460 nCi/ml, corresponding to a four fold rise in the tritium level It 
is not clear whether this is attributable to additional, fusion reactions occurring when the 
cell was quiescent or whether the earlier built up tritium, continued to leach but into the 
electrolyte- &**&+** 



% ’% 

The results of the neutron and tritium measurements carried. out during the above 
runs are tabulated in Table HI. The observed tritium concentrations have been corrected 
for enrichment effects due to electrolytic separation of deuterium and tritium as well as 
evaporation losses. It may be noted from this table that tritium production is much 
higher than the neutron yield, although in ‘hot fusion’, A hlir probability is known to be 
approximately equal. Our experimental observation jp that- both, neutron and tritium, 
generation seem to be occurring simultaneously feglkuse as mentioned earlier from the 
results of the the second run it is seen no neutron and tritium peak was observed for a 
long duration but the tritium level of . 43^* elt'Ctf oly te was found to have increased 
multifold after a neutron burst was noticed. The observation in B U N 2 of the electrolyzer 
shows that both neutron peaks and tritium were recorded only about 30 hours after the 
current had been put off. An important observation of this work is that "spent" Pd 
electrodes seem to lose their capability to support cold fusion reactions as can be seen 
from a comparison of the results of 'RUN 1 and RUN2. In the latter case the number of 
neutrons and tritium atoms produced has decreased This observation calls for further 
investigations A multi dimensional characterization of the freshly deuterated and spent 
rd electrodes, such as of the metaUographic and lattice structure will .go a 
long way in understanding, this phenomenon and nmv shed considerable light on the 
mechanism of ‘Cold Fusion’ 

Ackno w ! edgem ents. : 

The authors would like to acknowledge the unstinted cooperation received from 
several scientists in various aspects of these experiments. They are extremely thankful to 
each and every one of them. They include C.K. Pushpangathan, V.H . Path. A run Kumar 
and N.P Sethuram of Heavy Water Division, and R.K Rout and L.V. Kulkarm of 
Neutron Physics Division. Shn H.K. Sadhukhan, Head, Heavy Water Division has 
contributed immensely through fruitful discussions 


1 M. Pleischmann et al, J. Electranal . Chem. 261, 301-308 (1989). 

2 S.E. Jones et al, Nature 338, 737 (1989). 

3 ELHYGEN Hydrogen Generator (Mark V), Manufactured by Milton— Roy Company, 
Shannon Industrial Estate, Clare, Ireland. 

4 A, Shy am et al. This Report, Paper A4 (1989). 

5 T S.Murthv et al, This Report, Paper A§ (1989) 


Additional Remarks on Neutron Counts Data of Table II and Fig. 2 

(1) It is observed from the neutron counts data of Run No. 1 (21st April 1989) plotted, in 
fig 2 that while the BF 3 bank has recorded at least 9 clearly visible peaks, the plastic 
scintillator has missed out some of the smaller peaks. The reason for this is obvio usly 
higher background level of the plastics scintillator arising from its sensitivity to, gammas. 
I he smaller peaks have apparently got buried in the statistics of the background.'” 

(2) The ratio of the counts under the peaks in the two channels after subtracting: the 
background is found to vary between 1.3 and 2.8 which is considerably value of -67 
expected from their efficiencies determined using a Pu-Re tOi&foffr source. This 
discrepancy may be attributed to the following points, (a) The gBsipe scintillator being 
sensitive to the gammas of Pu— Be would indicate a higher sensitivity, (b) The energy 
spectrum, of cold fusion neutrons could be different from that pfTu— Be source neutrons 
A l fr CS t “ e ,, ener J y , res P° nse characteristics of the plastic acMiilator and BF 3 bank are 
different ^ effiaency rmtlos for ^ u ~ Be neutrons and cold fusion neutrons could be quite 

,#*% 4S# 
w 4> 

(3) From a detailed analysis of the neutron multiplicity spectrum (riven in paper A4) it 
r i nf B ^ 0unr ^ that between 10 to 25% of the cdtl fusion neutrons are emitted in bunches 

oi 100 or more. While the BF3 bank which, is surrounded by a moderator is able to 
resolve and count individually a number of simultaneously incident neutrons, the plastic 
scintillator will only give a single pulse fei though of a much larger height). In fact by 
comparing the measured ratios of the counts under the peaks in the two channels with the 
expected efficiently ratio one can approximately assess what, fraction of the neutrons 
emitted m each burst is due to bunched neutronic events. In the 21st April run it is 
founc, that 1.1 general the larger peaks seem to contain a higher fraction of bunched 

cv C-UbS . a 


A 1 


Details of the Milton Roy Electrolytic Cell 

Vol. of Pd cathode 
Mass of Pd 
Area of cathode 

Current density 


Volume of electrolyte 

Interval duration 
Efficiency of BF3 Bank 
Efficiency of ME 102 A 

7 cm 3 
82 gm 
300 cm 2 
60 amps 
£200 mA/cm 2 
20% NaOD in D 2 0 v ^ V 

V. - 


Neutron Counts Vs .Time 

+ 4 \ 

. %=S'mmutes 
' =0.06% 

% =0.4% 


Summary of Neutron and Tritium Yields 
(Current Density = 200 mA/em 2 ) 

Run No.& 


Duration of 




Atoms of 



Neutron to ’/“i / 
Tritium yield 
Ratio- .7.'“' 

Run 1 

2 1st April 89 

72 hrs 

4xlG 7 


Run 2 

12th— 16th June 89 

54 hrs 

9xl0 8 

oxlO 14 * 'I>^ l "4,8xl0“ 8 
1.9x1055* \ 0.5xl0‘ 8 

•These were measured 30 hrs and 27 
current had been switched off. 

days respectively after the 

% % 

„ V' 

P ? S r™?wnn^ , S;'E 0 i' C0LD FUSI0N STUDIES USING 

M.G. Nayar*, S.K. Mitra*, P 

M.S- Krisknant, S.K. Malhotratt 

D.G, GaonkarH, S.K. Sikka + A Shyam + and V. Chitra 

•Desalination Division, 

1 Heavy Water Division, 
^Neutron Physics Division. 


In their first cold fusion paper 1 Fleischraann et al suggested that mi electrolytic cell 

ff; V °}i mie f Urfar 1 area hi % h currerit density 'iSay cause fusion reactions 
resulting m the production of significant amounts of .heat and nuclear particles. The 
experiments reported m this paper present the rasuljg 'Of Pur early efforts to design and 
operate a high current modular Pd-Ni electrolytic cell and look for cold fusion reactions. 

Electrolyser and Operation .. «■ ^ 4 '%> 

A five module cell of bipolar filter pr.ess configuration having paIladium-{25%) 
silver alloy {procured from M/s Johnson-Mathey (U K )] as cathode fo.I mm thidkness 
and porous Nicke. as anode was fabricated. The electrodes are of circular plate geometry 
having a sectional area of ~ 78 cm?; The. five modules of the cell are connected m series 
de h °P erated even upto a high current density of 1 A/cm* and temperatures of 
, 1 j mixed products are carried out of the ceil and recombined in a 

buxnerpiondenser unit and the resultant heavy water recycled back to the electrolyser. 

snffirW if W i Cari I® r - arcu!at , ed T a co e, ler to kee P the operating temperature 
sufficiently low to reduce evaporation losses. Fig.l gives a schematic diagram of the 

modular cell. A flow diagram of the electrolyser and accessories is shown in Pig.2. 

The D| and 0 2 gases produced in the electrolyser and evolving out in a mixed 
s ream enters the recombination unit. This unit is essentially a burner complete with a self 

V°°n n6 * a ( rai )S ement 80 that the product of recombination namely D 2 0 is 
condensed and collected. There is a provision to add incremental quantities of oxygen 
continuously to ensure the complete conversion of D 2 to D 2 0, 

The system was filled with freshly prepared NaOD in D 2 0 (20%) upto the preset 
mark in the level gauge and the electrolyser switched ON and operated continuously at a 
current of 60 to 65 amps (corresponding to applied voltage of ~ 12.5V) from 5th May 1989 
onwards, with a few hours of interruption due to failure of the DC. power supply. 

Occasionally the current was raised to 78 A, The summary of characteristics of the cell are 
shown m Table I. 

The present operation was carried out to test the electrolyser, recombination, unit 
and other subsystems. Sustained operation for extended periods of time, particularly in 
ciosed loop mode, has not been carried out in. this preliminary study. 

Neutron and Tritium Measurements 

The placement of neutron detectors around the ceil was similar to that described m 
paper A- 1 I wo neutron detectors namely a bank of three BF $ detectors embedded m 

A 2 

employed to monitor the neutron output. The counts data was printed out on scroll 
printer. The BF 3 channel was counted for 110 seconds each, while the plastic detector 
counted for lOQsecs each. But both, counting intervals were commenced at the same time. 
The cell was operated continuously for 4 hours, when a big burst of neutrons overlapping, 
two counting intervals was recorded in both the detectors. Table 11 summarises the 

neutron counts data. 

The tritium, content of the electrolyte before and after electrolysiaias ! ' : daalyBed by 
liquid scintillation counting techniques (see paper A 9) were 0.055 nOi/ml and 190.8 
nCi/ml respectively. The significant quantities of tritium carried away: by the 

Ds and O 2 gas stream has not been accounted for in this, 
volume of the electrolyte in the system including hold ud was 
total tritium build up of 190 pCi or 4xl0 15 atoms 



Considering that the total 
1 litre , this corresponds to a 

if •• 



The electrolysis though carried out for limited periods only has shown conclusively 
that cold fusion occurs m this system also. This is obvious from the neutron bursts 
obtained (Table II) and the results of tritium .^n^yli's in the electrolyte. The latter results 
have shown an increase of over 3500 times in the electrolyte after 50 hours of operation. 

From the efficiency of the neutron dftectors and the neutron counts data given in 
i able II, it is inferred that about 5x1 0®- neutrons were produced over a 100 second interval. 
I he corresponding tritium yield was -4xlO i5 ; suggesting a gross neutron to tritium yield 
ratio of -10 9 . But it must be .emphasized that a considerable quantity of tritium may 
have been carried away by the gas stream. Although efforts were made to recombine and 
recover this D 2 0 for tritium it was not successful. The estimate of 10* 8 for the neutron to 
tritium yield ratio may therefore be considered as a lower limit only. 

The limited experimental studies conducted with this system has given enough 
experience for initiating a prolonged closed cycle operation, where simultaneous counting of 
neutrons and monitoring of progressive build up of tritium during electrolysis is proposed 
to be carried out. 



A number of scientists and engineers of the Desalination and Heavy Water 
Divisions have contributed in the fabrication, installation and operation of this cell. Their 
help is gratefully acknowledged. 



Summary of Characteristics of Five Module Cell 

A 2 

No, of Cathodes 

Area of each Cathode 

0:78 cm 2 per surface 

Volume of each Cathode 


crl cm 3 

Total volume of electrolyte 
in system 


„ "Tv 

1000 cm 3 p- 



60 to 65A with occasional 
peaking at 7$ A 

750 mA/crri* 

Current density 



Neutron Counts Variation 

BF 3 channel jv, 4 ’ 

(counts per 110 sees) \ 

Plastic Detector 
(counts per 100 secs) 

131 /> /V 


mi / > 


143 V 


m * 




■' 148 






* 125 






* The ratio of counts in the two channels were different in the two 
consecutive time intervals presumably because the single neutron burst, is 
partitioned unequally in the two consecutive intervals. This was because the 
counting intervals was longer by 10 seconds in the BF$ channel. 











dr* %■ 


.End Flange 












M.S. Krishna®*, S.K.Malhotra*, D.G.Gaonkar*, M.G.Nayart 
A.Shyamt and S.K.Sikkat 

♦Heavy Water Division, 
iNeutron Physics Division, 
tDesalination Division 


Since the two communications^ 2 reporting the occurrence of cold fusion 
experiments had been initiated in a number of laboratories to study the electrolysis of DjO 
with palladium (Pa) as cathode. In a few cases titanium (Ti) has also been used as cathode, 
li is a material of interest as it can form deuteride upto the composition of TiDg (against 
u.o m of rd). further Ti is more easily available and cheaper in our country. Three 

groups have reported the use of I i as the cathode material in their electrolytic 
experiments. Meanwhile in an interesting paper® .use of Ti in deuterium gas loading 
experiments has been reported wherein occurrence of neutron bursts under non — equilibrium 
conditions was observed. > \ * 

i j , obta ( ned ver Y encouraging neutron and tritium yields with electrolytically 

loaded Pd', it was decided to investigate Ti as cathode material in the electrolysis of D 2 0. 

°* ’Doth neutron and tritium measurements are reported in this paper. Unlike 
the i u— Mi cell described in paper A1 which was a diffusion type commercial hydrogen 
generator, the li electrolytic.- cell employed for this work was of very simple design and 
constructed from materials readily available. 

Description of the Cell', .*■ . 

A schematic. sketch of the cell is shown in Fig.l. A cylindrical rod of Ti of dia 22 
mm and length 150 mm constituted the cathode (area 104 cm 2 ) while an SS outer tube of 4 
cm ID served as the anode. The two electrodes are fixed coaxially with PTFE spacers and 
gaskets as shown in the figure. The inter electrode gap is 9 mm (surface to surface). As the 
cell does not have provision for separating deuterium and oxygen, both these gases are 
vented out from the top. The cell is also equipped with a thermowell to measure the 
temperature of the electrolyte. 5N solution oi sodium deuteroxide (NaOD) in heavy water 
(isotopic purity > 99.8%) was used as the electrolyte, to achieve high current density. An 
external DC power supply was employed. The salient features of this electrolyzer are 
summarised in Table I. 

Operation of the Cell 

electrolytic cell was first operated with 5M NaOD in D 2 0 for about 8 hours 
Only temperature increase was recorded for this run. The current was raised slowly from 
20 A (current density 193 mA/em 2 ) to a maxim mu of about 40 A (current density of -400 
mA/cnri). I he cell was then drained, flushed with distilled water several times, and 
operated for 12 hours with an electrolyte of 5 M N&OH in HgO, By this time the neutron 
monitors^ were mstaiied in position. Subsequently it was cleaned, rimed and a repeat ran 
with d m ISaOD m D 2 0 was conducted. As this cell does not have provision for automatic 
electrolyte feed, the electrolysis was earned out as a batch process. After a few hours of 
operation, when the electrolyte level had decreased considerably, fresh electrolyte was 
added to make up the volume. The cell had operated for 8 hours, when the cathode was 
found to have attained a dull black coating which when analysed chemically was found to 


A 3 

contain iron as major component. The electrolyte solution also developed a pale greenish 
yellow colour because of dissolution of Fe from the anode. The cell is therefore being 
modified to have both anode and cathode of titanium In addition the inter electrode 
distance between anode and cathode is being reduced to attain larger current densities. 

Neutron Measurements 

A bank of three He 3 counters embedded in paraffin moderator and a plastic 
scintillation detector were employed for neutron measurements. Eftbd&ncy^ of the lie 3 
detector viewing the cell was ,1%. The plastic scintillator which was placed away from the 
cell was employed as a background monitor. Background data collected before 

electrolysis was commenced. The background counts m the He 3 ehaime: was ~ 240 counts 
per 10 seconds. Neutron measurements carried out with the cell operating with NaOB-HgO 
as electrolyte was similar to the background data. But when itie cell was operated with 
N'aOD-D'jO as electrolyte, after about 3 hours the level of fife neutron counts was higher 
(- 590 counts per 10 secs.). No big bursts of neutrons observed as in the case of Pd— Ni 
electrolyse? 7 . When higher counts was continuously jj4fwd. the cell was put off and the 
neutron counts was observed to come down irnmedl.^e1y; but it was still above background 
{.-3-85 counts per 10 secs,). The experiment was .finally terminated when the counts level 
reverted to background presumably due to fofihiig of cathode surface by iron deposit. Fig.2 
shows a representative plot of tne data ffortJOlsi'in before and 5 nun after switching off) 
showing the typical drop m the neutron counts, when the cell was switched off. The typical 
background counts when the ceil wail, removed is also indicated in the figure for 

Tritium Measurements 

At the end of about 6 hours of electrolysis when the electrolyte was sampled and 
analysed for tritium content it -was found to contain -48 nCi/ml of tritium activity. This is 
almost three orders of magnitude higher than the tritium level of the stock heavy water 
used for preparing the electrolytic solution which was -0.05 nCi/mi. The tritium level in 
the electrolyte after .about 12 hours of operation with NaOH was only 0.0676 nCi/ml. 

Conclusions ‘ ^ 


The results of the neutron and tritium measurements seem to strongly indicate that 
cold fusion occurs in the Ti— D system also. When compared with Pd— D systems where big 
neutron bursts are observed, in the present system the neutrons are produced at a low but 
more or less steady rate. It is estimated from the neutron counts data, the duration of cell 
operation (the integrated time for which the cell was active was approximately 150 
minutes) and the efficiency of neutron detectors (1%) that in all about -3 10 7 neutrons had. 
been generated. The terminal tritium level m the electrolyte was -48 nCi/ml. Since the cell 
volume is -150 ml the total tritium inventory m the cell corresponds to 48x150 = 7.2 fiCi. 
After correcting for the tritium initially present m the D 2 0 both m the initial charge as 
well as make up volumes, it is estimated that an excess of approximately 7 pt Ci of tritium 
was produced due to cold fusion reactions. Considering that the total number of neutrons 
generated was -2 10 7 , this leads to a neutron to tritium yield ratio of approximately -10 -7 . 


The authors are very thankful to Shn S.K. Mifcra, and Shri V. Raghunafhan of 
Desalination Division arm hum v , Cmtra of Neutron Physics Division for their help in the 
experimental work. We are also -grateful to Dr. M Srimvasan, Shri H.K. Sadhukhan, Dr. 
Mi .5. Ram am and Dr.P.K Iyengar for their keen, interest in this work 



M Pleischmann and S Pons, J, Eiectroanal. Cbem., 261 
bE Jones et al, Nature (London) 338, 737 (1989). 

G K Mathews et al, Indian 3. Tech. 27, 229 (1989). 
h 5 V S&nth&n&m et al, Indian J. Tech. 27 175 (1989) 
G banchez et al to be published m Solid State Comm. 
A Ue Nino et al, Europhysics Letters Q, 221 (1989) 

M.b. Knshnan et al, This report, Paper Al (1989) 



Salient Features of the Ti-SS Cell 

Volume of Ti Cathode 
Area of Cathode 

Current Density 

Volume Of Electrolyte 

r :>V 

57 cm 3 

104 cm 2 

40 A- 60 A 

< 600 mA/cm 2 

20 % (5 Molar) NaOD in D 2 O 

150 cm 3 



A 4 


A. Shyam, M. Srmivasan, S B- Degwekar + and L.V. Kulkarm 

Neutron Physics Division 
+ React or Analysis and Systems Division 


Since the first announcement by Fleischmann k Pons ! ar|8* Shortly thereafter by- 
Jones efe al 2 of the observation of cold fusion reactions in palladium electrolytically loaded 
with deuterium, various theories and speculations have been pul forward as possible 
mechanisms for the same. All the schemes proposed so far may be classified into two broad 
categories : (i) Those that lead to (d— d) fusion reactiorjS-.takihg place one at a time Le, 
wherein occurrence of one fusion reaction does not directly influence the probability of 
occurrence of another. In this case one can assign a cer lJp probability per second per (d-d) 
pair for the reaction rate, A figure of lflf* has fqp^xkpi'pi* been deduced for this by Jones 
ft al. (ii) The second, category of mechanisms deads To a cascade or sharp bursts of fusion 
reactions. /L 

One of the earliest speculations 3 ' 4 attributed the cold fusion phenomenon to muon 
catalysis triggered by cosmic ray produced muons. It was pointed out that, each muon could 
in principle catalyze several hundred fusion reactions within a time span of a couple of 
microseconds. Yet another mechanism proposed has been lattice crystal fracture or 
cracking leading to acceleration ^deuterons to energies of 20 to SO Kev in electric fields 
generated -across fracture crevices Such internally accelerated deuteron beams are then 
presumed to cause (d-d| reactions. Here again the fusion reactions may be expected to 
occur erratically leading thy bunched occurrence of (d— d) reactions. A brief review of the 
phenomenology associated with lattice cracking is presented by Kaushik et al in Ref. 5, 

Since neutrons are one of the end products of (d — d) fusion reactions, it may be 
expected that statistical analysis such as measurement of the multiplicity spectrum of 
neutron emission can give valuable insight into the possible origin of cold fusion reactions. 
The emission M neutrons in bunches of two or more can for example be observed by 
employing two (or _ more) fast neutron detectors and looking for coincidences among 
detected, pulses within agate interval of say 1 ear 10 #*$ Since the average background count 
rate is generally small, the chance coincidence rate due to background events would be 

negligible, making the task of establishing the occurrence of multiple neutron emission 
events quite easy. 

An alternate technique of delecting fast neutron multiplicity is to employ a thermal 
neutron detector surrounded by a hydrogenous moderator such as paraffin. This type of 
detector system has the interesting property that a bunch of fast neutrons simultaneously 
incident on it would be temporally separated due to the statistical nature of the neutron 
slowing process and get detected as individual neutrons within a time span governed by the 
neutron die-away time in the moderator-detector assembly. It is this type of thermal 
neutron detector that was employed in the present studies. This paper summarises the 
results of the neutron multiplicity spectrum measurements carried out with the Milton Roy 
commercial electrolytic cell® as well as- Dg gas loaded Ti targets 2 "*^. 


Theory of Multiplicity Analysts 

A 4 

tf the events producing neutrons are. random in time and result in one neutron par' 
event, then the number of counts observed in a time interval r would be distributed 
according to the Poisson law as follows: 

where P, is the probability of obtaining r counts in time r and No is the average count 

rate Thus when the average count rate is small i.e. when Nor < < 1, one can set e~^° r a 
1 in this case the probabilities of detecting one, two and, three neutrons in the time 
interval r are (Nqt ), (Nat) 2 /2 and (N 0 r) 3 /3 respectively, Irfparficular note that the ratio 
of doubles to singles is (Nqt/ 2) while that of triples to doubles is (Nor/3) and so on. 

On the other hand if there are events which/ in bunches of neutrons, say v 
neutrons per bunch, then in a time interval r whiSk /encompasses the nuclear event, the 
probability distribution (P r ) of counts would a binomial distribution as follows: 

Pr “ « r (2) 

It is presumed that the interval f is large compared to the neutron die-away time in 
the detection system. It is also assurped that the source event rate is so low that only one 
such event occurs m interval r. Here e is the overall counting efficiency. If S is the event 
rate of such multiple neutron produfeingxascades its contribution to the average count rate 
would be Si/e. - j % ‘ 

Thus we can now conceive of a situation wherein there is a steady random 
background count rat|#Plfo c.p.s, superimposed on which is a signal event rate of S events 
producing v neutrons each In such a case the average count rate in time r can be expressed 

as: '%# r 

^ N r + StvT (3) 

Likewise the probability of registering two and three counts in time r now comprises 
of two components: One due to random background events and the second due to signal 
events releasing u neutrons each. Table I summarises the expressions for these probabilities. 
Note that while the probabilities of various multiplicities due to the random background 
depend on the product (Nor ) that due to the bunched neutronic events depend mainly on i> 
and t. In the limit of v >> 1 and e « 1 the expressions simplify as shown in the last 
column. It is dear from this that the deciding quantity under these circumstances is the 
product Vi, 

Table II gives the expected frequency distributions of multiple neutron, counts for 
typical values of No, t and v. The counting time interval (r) is kept fixed at 20 ms while 
the bunched neutron producing event rate (S) is taken as 10~ 2 / s - The data presented is for 
a total of 10 s sampling time intervals. It is clear from, the table that while the average 
count rate for Poisson events is much higher than for bunched events, the frequencies of 
higher count multiplicities ( >2) is much higher for the bunched events. 

Experimental Details 

Neutrons were counted by a bank of thermal neutron detectors embedded in a 
paraffin moderator block, One bank comprised of three BPg counters while the other was 




A 4 

made up of three He 3 counters. The neutron die-away time in each of these was -25 ps. 
The BF3 bank was mounted close to the Milton Roy Pd-Ni electrolytic cell and the He 3 
bank near the D 2 gas loading apparatus about 1.5 m away. While one counter bank was 
being used as signal counter in one experiment the other bank served as background 
monitor and vice versa. The efficiency (e) of detection for the cold fusion source neutrons 
was typically in the range of 0.5% to 1.5% depending on the exact distan ce between the 
cold fusion source and the detector assembly as well as the pulse height discriminator bias 

The outputs of both these counter banks were fed to scalers whose readings could be 
read off by a Personal Computer (PC) at the end of each counting interval which was 
controlled by the clock in the PC. By taking the difference!' Trf%ifie scaler readings 
corresponding to the end of two consecutive counting intervals: the number of counts 
recorded in a given counting interval is computed and stored the PC. It took the PC 
typically over 250 ms to carry out these operations following each sampling time. Hence a 
set of 1000 samples consumed a real time of -5 minutes. '* 

The duration of the counting interval selected was -20 ms. This is a compromise 
between the conflicting requirements of increasing the total number of counts accumulated 
to get good statistics and minimizing the chance coincidence probability for Poisson events. 
For the typical count rates met with during the neutron emission phase of the cold fusion 
experiments 20 ms was a reasonably satisfactory choice. From such data accumulated over 
several hours, the frequency distribution of counts recorded in 20 ms interval could be 

Results and Discussion 


Fable III summax|$e$f l$|Kvarious neutron multiplicity spectrum measurements 
carried out during May-July ’89 m cold fusion experiments. The details of the Milton Roy 
commercial electrolyser and its operation are described in Ref. 6. Likewise the D 2 gas 
loading experiments ar% summarised in References 7 and 8. To begin with the statistics of 
background counts was studied to ensure that the equipment was functioning satisfactorily. 
For this purpose all potential cold fusion sources were removed from the room where the 
neutron detectors were located. Data acquisition of background counts continued in an 
uninterrupted manner for 63 hours over a week end (1800 hrs on Friday 2nd June to 0900 
hrs Monday 5th June). During this run the average background count rate in the BFj bank 
was -0.023 cps and in the He 3 bank -0 43 cps (The background rate in the BFg bank was 
intentionally adjusted to be lower by setting its pulse height discriminator level high.). 
Table IV presents the results of the frequency distribution of counts obtained in this long 
background run. It is heartening to note that not even once out of the -750,000 odd 
samples were 3 or more counts registered by either of the detector banks as may be 
expected on the basis of Poisson distribution. The ratio of doubles to singles frequency 
further conforms to Poisson statistics, indicating that the equipment was functioning 
properly. Sparking in any of the counters for example would have given rise to significant 
non— Poisson behaviour. 

Table V represents the results of our first attempt to measure multiplicity 
distributions of neutron signals. The data was accumulated overnight (1805 hrs on 26th 
May to 0645 hrs on 27th May) with the BF 3 bank viewing the Milton Roy cell which was 
quiescent i.e. the cell current was not on. Besides, a plastic scintillator (NE 1Q2A) biased 
to register only neutrons of energy >9MeV monitored cosmic ray and other background 
events. The first column of the table gives the probability distribution of counts of the BF3 
bank for 20 ms intervals. The average count rate works out to 5-6 c.p.s. It is clear from the 
frequencies of 2s, 3s and higher multiplicities that there is considerable contribution of 


A 4 

non— Poisson events. The second column of the table gives the frequency distribution of the 
same counts data whenever there was a pulse recorded by the plastic scintillator also 
during a 20 ms interval. From this we conclude that only about 1% of the neutron events 
occurring in the BF 3 bank can be attributed to cosmic ray showers. 

Table VI presents the frequency distribution results of the Milton Roy .cell run of 
12th to 14th June. As noted in the companion paper 8 six neutron bursts ,d|j4)ve minutes 
duration each were recorded during this period, the first about 50 miiSA after cell 
electrolysis commenced on 12th June, the second and third about an hour thereafter and 
the remaining three a few hours after the cell current was switched off on ihe evening of 
14th June. During the burst phase, the count rates were in the range of -0.5 to 1.7 c.p.s. 
which is about 4 to 14 times that of the background value (*.0.32 c p s.). However it is 
noteworthy that in 4 out of the 6 bursts observed, count mul^gplicifies of 2, $, 4, 5 and even 
10 have been recorded at least once each. This type of-beh^vibur is indicative of high 
multiplicity neutron emission events. Throughout thilwTu'fi lasting several days the 
background counter did not record any noticeable increase of counts. 

Table VII summarises the frequency distrib|!Siqif measured during the 2.5 hour long 
neutron burst recorded on 16th June from 1900 hours onwards with the Milton Roy 
electrolyser (Fig. 3 ol paper A.l gives a plq>t of the count rate variation during this burst). 
It may be noted that the cell had not beep operated for the previous -52 hours. The count 
rate during this wide neutron burst attained a value as high as 20 c.p s. at the peak. It 
rnay be noted that the background neutron monitor which was only 1.5 m away also 
indicated a small increase in count rate confirming that the neutrons originated from the 
electrolytic cell. Careful scrutiny of these results indicates that the frequency distribution 
essentially corresponds to a Poisson diatrbution. However, the fact that, multiplicities of 5 
or more are recorded. severafiMtes again points to the sporadic occurrence of multiple 
neutron emission events. Iyichofeworthy that around 1950 hours (close to the peak) there 
were more than 20 sup. multiplicity cascade events within a time span of 5 minutes. 
(Please note that in Table A?1I set M corresponds to the early part of the burst and set A 
close to the peak with absolute time increasing from M to A). 

Tables V HI Jr IX summarize the results of multiplicity distribution measurements 
carried out with two D 2 gas loaded Ti targets. During the weekend run of 9th to 11th June 
with 15 gma..dfefTi— Zr deuteride (see Table VIII) the average count .rate measured was only 
0.42 c.p.s. ‘Since this corresponds to an NqT value of 0.008, we expect a doubles to singles 
ratio of 0.004 only. While the high doubles events in both the background and signal 
counter can possibly be attributed to statistics, the 3s and 4s in the signal counts can be 
attributed only to high multiplicity neutron emissions from the deuterated Ti-Zr target. 
Absence of such high multiplicity events in the background channel further strengthens this 

, .Table IX presents a similar result from a Ti disc target. The gas loading procedure 
for this target is described in Ref. 7, while the measurement of spatial distribution of 
tritium on the surface of this target by autoradiography is discussed in Ref. 8. The neutron 
active phase of this target lasted almost 85 minutes during which it is estimated to have 
emitted -5 x 10 s neutrons in all (see Fig. 1). On the whole this target also points to the 
occurrence of a significant number of high multiplicity neutron emission events. 

The last columns of Tables VI, VII and IX give the total frequency distribution for 
all the 5 minute duration i.e the sum of all the columns. Also shown are the theoretical 
frequency distributions, expected on the basis of Poisson statistics. It is worth noting that 
except for background case (Table IV), in all the cold fusion measurements the 
frequencies fall according to the Poisson law for low multiplicity events but there is distinct 


tendency for them, to show a slight peak between the multiplicities of 4 and 6. If we 
assume that this peak is due to the superposition of bunched neutronic events on a Poisson 
background, we deduce the value of v to be m the range of 400 to 600 since the peak of the 
binomial, distribution occurs at the multiplicity value of m (see Table I) and e values in the 
experiment are typically -0.01. Further, by comparing, the observed frequencies with those 
shown in iable II, it appears that the average source event rate for such events during the 
neutron emitting phase is very roughly about 10~ 2 per second 

Jr "y 

Summary and Conclusions 

The multiplicity spectrum of neutron counts observed in 20 ms time intervals has 
been measured so far six times during cold fusion investigations, pfihwith electrolvtically 
Joaded rd electrodes, two with gas loaded Ti targets and -.cneeAsith the background. 
” the background counts displays strictly Poisson statfffci^in the three cases where 
distinct excess over background neutron bursts were recorded~between 10% and 25% of the 
emitted neutrons appear to display high neutron multiplicity characteristics. The observed 
frequency distributions can be explained as being .. d®- to bunched neutronic events 
superposed over a Poissoman background Analysis M Jhe data leads us to the conclusion 
that these bunched neutronic events occur on .^'average once every hundred seconds 
during the active period and typically produ^boift 400-600 neutrons per bunch, within a 
tune span of less than 20 ms. Such occasion al%Iut.ron bunches are also reported to have 
been observed by Menlov et al at LAMIA % 

this is viewed in the light, of the experimentally deduced neutron— to— tritium 
branching ratio of 10' 8 , we are oj^gefktl come to the intriguing conclusion that during 
bunched neutronic events upwards of 1CM® fusion reactions occur in under 20 ms®. As this 
appears very unlikely, the authors are inclined to believe that bunched neutronic events are 
not accompanied by tritium production with a yield ratio of 10* and hence lattice 
cracking 0 where the brapMepratio is close to unity would appear to be the most plausible 
cause lor bunched neutrorffieprission. 


1 FRisehmaan and S. Pons, "Electrochemical ly Induced Nuclear Fusion of 
Deuterium , J. Electroanal. Chem. 261, 301 (1989). 

2 S.E. JoneWf al, "Observation of Cold Nuclear Fusion in Condensed Matter", Nature 
338, 737 (1989). 

3 5 U i na f ! ? ? A Cha Pjf ne and R * W - Moin, "Catalysis of Deuterium Fusion in 
Metal Hydrides by Cosmic Ray Muons", UCRL preprint 100881. 

4 Report on the Workshop on Cold Fusion held at §anta Fe, New Mexio, USA held 
during May 1989, Science News, June 3 (1989). 

5 T.C. Kaitshik, M. Srimvasan and A. Shyam, "Fracture Phenomena in Crystalline 
Sohds: A brief Review in the Context of Cold Fusion", This Report, Paper 05 (1989). 

6 M.S, Knshnan et al.., "Cold Fusion Experiments using a commercial Pd-Ni 
Electrolysed*, This Report, Paper Al (1989). 

7 V.K. Shrikande and. K.C. Mittal, "Deuteration of Machined Titanium Targets for Cold 

b 1 nrt * * HAL C _ O JL TV f ^ 

usion txpenmenis 

is Report, Paper B2 (1989). 

This Report, Paper B3 (1989). 

P.1C Iyengar, "BARC Cold Fusion Experiments", Paper presented at the 
Goaf, on Emerging Mud. Energy systems, July 1989, Karlsruhe (FRG). 

Fifth lat. 




Total counting time 

Random count rate 

Counting interval 

Detection efficiency 

Average No, of neutrons per bunch 

Bunched Neufcronic event rate 

of counts 


Random Events 

val u es for 

Nq T< < 1 %, 

Bunched Events 

values for e<<! 


I e -No, 

| (1— Not) 

ST e~ U£ 


|(N 0 r)«- N » T / ' 

| (Nor) 


A % 

n % -d r 

X %) e - N « r 


ST(|)(l-e) i/ ~ 2 e 2 


uj 4. t 

T^ 3 e -N 0 r 


ST(3)(1— e) ,/_3 e 3 

ST^ 3 < 

T (N»r) 4 -N 0 t 

r 4! e 


ST^Xl-ef" 4 ^ 

ST^ e 

T(N|ri 6 e -N„r 


ST(5)(l-c) i/ ~ 5 e 5 

, h5 

ST^Jf e 






Expected Frequency Distribution of Counts for Baade* aad 
Bunched Neutronic Events for Typical Sets of Parameters 





Frequency of counts 

in 20ms inter 

vaJs for HP samples (8hrs real 

Lme}#L . 

Poisson Events 

Bunched Events (S =!0~ ! per seel ~ 


N c =3 0 CPS 

v - 100 
< = 0 .016 

Sjȣ = 0.C45 

* = 100 
f = 0.005 
Svt = 0 003 

* = 5 U0 
i = 0 015 

See = tJfSeL 

v = 500 
" ~ < = 0 005 

Sj/« = 0.025 












6 1 

0 07 V/ 

4 0 


-TCT 2 






<if s 

<10~ 2 







-I0* 4 






-10-* 3 



0.003 f>.- 




Counting interval 20 ms 

Total counting time 63 hrs 



Lreqiieacy Distribution, of Counts in BF3 Bank and Plastic 
Scintillator with Quiescent Milton Hoy Cell 

Counting period 
Counting interval 
Total number of sampling 

20 ms 
144 Q0&. 

of counts 

Gross Frequency 

Frequency in those 
samples in which 
piffle 'sc 1 nt ill ator 
r ffecords a pulse 
Of large height 




































Frequency Distribution of Counts for 1000 Sampling Intervals Each of 20 as 
During Six Periods of High Neutron Activity 
(Milton Roy Cel 1 Run of 12th to 14th June 89) , 

city of 

Frequency Total (A to F) 

12th June ] 4th June 

A B C D E F 

\ (Observed) (Expected) 






















3 1 







| o 


<1Q* 3 





lf 0 






1 * 

4 % * 







jr * 

; .^\o 







’ 0 







o 0 








. :0 








■ "0 







Frequency Distribution of Counts for 1000 Sailing intervals of 20 m Each 
During Periods of High Neutron Activity 
(Milton Roy Cell Run of 16th June) 


Counting Interval 

16th June 
20 m» 

city of 


Total (B toM) 


G H 







^ Observed 














S'V 23 






7 VT* 







2 \ 







0 1 

,4 * 













0 •' 

O' ,? 














Q j 

0 ' 







:"0 ,« 

0 ^ 






-lO* 7 































-10-' 3 




























































24 345 

7 9S 

3 1.6 

3 2 

355 104 447 492 

33 42 

4 2 




















0 0 
0 0 

fig ^ 


t> 0 













.. % 







# i 




oa a Deuterated Diac for 1000 Sampling Intervals of 20 «as Each 
During Periods of Bigb Neutronic Activity 

r ' 

city of 



Total {A to Q) 
















Observed Expected 

























' & 




















% 1 















































































T.P.RMhakrishnan, R Sundaresan, J, Arunachalam, 
V.SitaxamaRajUjR.Kalyanaiaxnan SGangadharan 
P.K. Iyengar* 

Analytical Chemistry Division 


hc % 

Recently, Fleischmann and Pons' have obtained evidencsCwtlhe fusion of deuterons 
induced at low temperature, through the electrolysis of deuterium oxide at a palladium 
cathode It is believed that, this can be achieved if the deuterium loading of the lattice 
exceeds the value for PdDo.g. In view of the relevance of this problem to this Centre, a 
program of work was initiated in this Division. Thi&rrepart is a summary of the behaviour 
of Pd and other metals during charging in alkahne,,he«yy water. 


Materials, Reagents and Apparatus 

w '^k, ^ 

Electrodes of different metals were used. These include a hollow Pd cylinder (area 
5 87 cm 2 , thickness 0.4 mm), Pd ring (area 14.5cm 2 , thickness 2 mm), Pd foil (area 1.5 
cm 2 , thickness 0 3 mm), Ti plate (area 8 cm 2 , thickness 0.5 mm) and a triangular piece of 
Ni— Ti alloy of area 10 cm 2 apd“ wjdght 10.5 gm. Platinum gauze, Pt disc or Pt foil of large 
area was used as the anode Heavy water (99.87% purity) and pure Li metal were used to 
obtain 0.1 M LiOD I alar— 2 ' N 2 or argon was used for deoxygenation and stirring. All 
electrodes were lightly abraded with emery, rinsed with acetone and dried before use. 

Electrolysis cells were made of quartz for the container and the lids were of PTFE 
or glass. Provisions and inlets for de oxygenation, addition and removal of solution and 
temperature measurement were made. Loss of D 2 O due to evaporation and electrolysis was 

compensated, ‘ 


Galvanostatic sources and current pulsing units were made in the laboratory. A 
Digital Panel Meter (PLA, DM— 20), was used for voltage measurements. Neutron counting 
and tritium activity studies were carried out, as described later. Set— up is shown in Fig. J . 


Differential Enthalpimetnc Studies 

A 5 

increased and attained a high value. The temperature rise in Cell I (Pd— Pt) is due to 
absorption, dissolution and interaction of deuterons in Pd lattice. 

Calorimetric Measurements of Enthalpy changes 

A modified isoperibol solution calorimeter with an accurate thermistor bridge was 
used. The cell with contents was enclosed in a tight-fitting Dewar and was immersed in a 
thermostat at 25°C for thermal insulation. Electrodes of hollow cylindrical; Pd cathode(5.48 
cm^) and Pt sheet anode with 50 ml 0.1 N LiOD were used Iolar— 2 argon was used for 
tie oxygenation and the solution was stirred mechanically. Charging was effected at different 
current densities and the terminal voltage in each case was noted. The solution 
temperature was monitored by the thermistor probe and bridge jakd also by a sensitive 
thermometer. The temperature increase as a function of tjme at different e.d, is shown in 
Fig, 3. The heat capacity of. the cell and components wap determined by electrical 
calibration and corrections for heat losses were applied ori the basis of Newton's law of 
cooling (Fig. 4). The results are included in Table 1 . ; ' 

Extended Electrolysis with Current Pulsing 
Cylinder Electrode : # 


Nk % 

A hollow cylindrical Pd cathode (5 9 cm 2 )and a Pt. gauze anode were used and 0.1 N 
LiOD was electrolysed in a quartz cell with nitrogen bubbling. Initially, a constant current 
of 1A was used and when the temperature reached 60^C pulsing was commenced between 1 
to 2 A at 1 s. interval. The temperature was controlled to 63°C by forced circulation of air. 
Neutron flux measurements were made and D 2 0 was added for make up. After 41. 8 hours 
during which 52.2 AH were .consumed, the electrolysis was terminated. Measurement of 
tritium build-up in theJnsf, solution showed 3,75 /iCi. 

Ring Electrode : 

W 65 ml ol 0.1 N LiOD in D 2 0 were electrolysed in a quartz- cell using a Pd— ring 
cathode (14 5 cm 2 ) and two Pt discs as anode. At low current densities, the cell voltage 
was observed as a function of current, and this is shown in Fig. 5. Later, electrolysis was 
carried out Tor nine days with current pulsing between 1A and 2A (80 h), 3 A and 4 A (20 
h)and between 4A and 4.5 A (7 hj consuming a total of 296 Amp. hours. Neutron activity 
and capture 7 measurements were made throughout the duration of electrolysis. At the 
conclusion of the electrolysis, tritium content in the solution was found to be 16.25 /i.Ci . 

( 2 ) The above experiment was repeated using 0.1 N NaOD and a total of 231 Amp. hour 
were passed during 77 hours. Neutron counting and tritium activity measurements were 
carried out. 

(3) The same ring electrode, after degassing in vacuum, was subjected to 844amp. hour 
charging in 382 h. in 65 ml of deaerated 0.1 N LiOD. There was no significant tritium 
build— up m solution. The electrode was greyish black in colour. 

Ni— Ti Electrode; 

A Ni-Ti cathode (10 cm 2 ) was used in 0.1 N LiOD and a total charge of 135 amp. 
hr were passed over ill hrs. Appreciable neutron activity was observed. The cathode was 
observed to flake off and disintegrate in powder form. 


A 5 


A Ti sheet cathode (8 cm 3 .area) and Ft sheet anode were used in. the electrolysis of 
1 ^ ' m ®20 at 34 mA.em -2 . The terminal voltage was 4,4 V and temperature rose 
from 2 7®Q to 36,6°C in 53 .minutes. Prolonged electrolysis did not show any significant 
increase in temperature. 

Nuclear Measurements: 

,v^S «K# 
* hp,<#' t§ 

Four different types of measurements were made to identify the emission of 
neutrons from the electrolysis cells. 

3 Detection and direct measurement of neutrons were based on the use of He 3 detectors 
arranged in a well— counter as weli as a Li 6 enriched scintillation detector (Bicron, Model 
NP— 2). The detectors were calibrated by Am— Be source Near complete rejection of high 
energy y-rays was ensured by proper discrimination of the 2 5 MeV Co 63 sum peak using a 

Co 60 source, .^5 

2. Detection of high energy capture 7 -rays of GdJPIvSid Pd was achieved by using Ge(Li) 
or HPGe detectors. “ J ' 

24 'V” 

3, Measurement of low energy capture 7 -rays Af energies 199, 944and 1186 KeV was carried 
out with a HPGe detector. 

4 Gross counting of 7 -rays of energy greater than 3 MeV was effected by means of a 
properly shielded 3" x 3" Nal (Tl ; ) . detector. 

| § Jp e| 

Ail the above measurements were done in combination to yield cross— validated 
results. However, in somsf celid,>only a single type of measurement could be made. 

Neutron count mg was performed both in the MCS mode (using 1BM-PC based 8 K 
MCA system) and m the PHA mode whenever possible. Different dwell times ranging from 
0.5 sec. to 60 sec. were selected in the MGS mode to check whether neutron emission is 
continuous or in "bursts". The complete counting set up is given in Fig. 6 . The results can 
be summarised as: under: 

a) Palladium hollow'- cylinder cathode (run 3,1) 

The 1186 KeV gamma ray activity was measured every 100 sec. interval for more 
than 24 hours. In Fig.?, three definite 'spikes’ can be identified. The duration of the 
‘bursts’ was 14 to 20 minutes. 

b) Palladium ring cel! (run 3.2.1} 

A time correlated analysis of the neutron counts in the Li 6 scintillation counter and 
measurement of 7-rays of energy greater than 3 MeV in the Nal (Tl) detector were carried 
out (Pig, 8). The correlation coefficient for the 50 observations was 0.28 which is significant 
at the 5% level. 

c) Ni— Ti cathode ceil (run 3.3) 

Neutron counting was earned out using He 3 as well as Li 6 scintillation detectors 
with the former in both the MCS and integral modes and the latter in the MCS mode. The 
cell was run initially at 1A current and when the current was increased to 2 — 2.5 A, three 
well— defined "spikes" were registered in the MCS mode as shown in Fig. 9. 


A 5 

The count rates of He 3 and Li e scintillator are not correlated in any significant 
manner, possibly because of the different efficiencies and geometries. However, tiny spikes 
are registered in the He 3 detector. 

d) Palladium ring ceil (run 3.2.3) 

Several spectra for the low energy capture gamma rays were detected using HPGe 
detector for durations from 10,000 to 50,000 seconds. In some spectra, 199, 944 and 1186 
KeV capture 7-rays could be identified though they were not traceable ini^&F%pectm. As 
the accumulated count rates for these energies are rather small, it ap^iSpr that neutron 
emission is low and. not continuous. % 



Electrolysis of alkaline heavy water results in the splitting of D2O with evolution of 
deuterium at the cathode and oxygen at the anode. The electrode process involves charge 
transfer, adsorption of the intermediates, subsequent reaction and gas evolution(2) On Pd 

potential an electrochemical desorption reaction is favoured. In the case of Ni-Ti, the 
discharge reaction is always accompanied hydrate determining electrochemical desorption 
reaction. jt \\ 

The uptake of deuterium by metals during charging depends on the reaction 

MD a< j s *— •—-# — M.D|a,ttice (1) 

and is, therefore, governed by gfe/fraction. of the electrode surface covered by adsorbed D, 
solubility of deuterium in tbeimetai and its diffusivity. The observed rise in temperature of 
the electrolyte is not tile exothermic dissolution of deuterium in Pd, Ni— Ti and Ti 

or due to other chemical; factors. 

The measured: overpotential is composed of ohmic, activation, adsorption, diffusion 
and concentration Over potential terms. The ohmic resistance and polarisation resistance 
cause Joule heating and thereby contribute to observed changes in enthalpy. Had the cell 
systems been reversible, then if the polarisation is reversed, the electrode reactions at Pd 
and Pt elec%pdes should also be reversed. In the upper limit the maximum Joule heating is 
the product of the cell voltage E and cell current i. In the case of Pd-Pt electrodes, the 
minimum voltage or back E.M.F for electrolysis of DgO was calculated from 
thermodynamic data as 1.54 V. Therefore, the electrical power available for Joule heating 
is f E— 1.54)i and this is the lower limit. The enthalpy changes calculated for different c.d. 
and the ratio of thermal output to the power input expressed as per cent breakeven are also 
included m Table 1 

time using a precision integral flow— meter and by gas chromatographic .analysis of the 
composition of the gas mixture. As excess heat is liberated over and above electrolysis, it is 
clear that some other reactions axe responsible for the excess enthalpy observed. Pauling 3 
ascribed the excess heat to the formation of palladium deuteride. However, Bockris and 
coworkers 4 have shown that exothermic effects due to solution of D m Pd, recombination of 
D atoms, formation of D2O etc. cannot, account for the observed heat evolution. Normal 

chemical reactions cannot account for the generation of neutrons or the production of 
tritium during charging of Pd with deuterona. As has been pointed out by Pleischmann and 
Pons 1 , the results can be rationalised and understood on the basis of cold fusion reactions 
occurring between deuterons in the Pd lattice as indicated below: 

2 2 

1 D + r D 

2 2 

t D + t D 

3 1 

1 H + , H 

3 1 

2 He + 0 n 

In the present work, the emission of neutrons, reasonably above the background 
level, and the build up of significant tritium activity in excess of the blank value, have been 
confirmed in four different electrolysis experiments. In certain, experiments neither the 
evidence for significant neutron emission nor any appreciab^B^ild 'up of tritium activity 
has been observed. It is likely that in such cases charging Was not sufficient for ensuring 
optimum loading of the lattice with deuterons for inducing fusion. However, it appears that 
in addition to reaction channels (2) and (3), the possible; occurrence of a non— emitting 
nuclear process cannot be precluded. This reaction carfpfe*. Written as 

2 2 lattice 4 * # « 

Pd +i D +i D 2 He + Pd / % 

catalysed # % 

'it % 

which implies that the lattice is excited to a higher energy level to conserve both 
momentum and energy. It is likely that, during the subsequent lattice relaxation, the excess 
energy stored in the lattice is liberated as heat(5) This mechanism would lead to the 
formation and build up of He inside the metal and can possibly account for the observed 
low yield of neutrons or tritium in certain experiments. Accurate mass spectroscopic 
analysis of He 3 /He 4 ratio jn the cathode material is needed to substantiate this view. In 
conclusion, it is necessary to examine in detail the different parameters and optimize 
important factors like metallurgical history and pretreatment of the cathode, solution 
chemistry, surface chemistry and electrochemistry to achieve reproducible fusion through 
electrochemical charging; of metals in heavy water. 


Authors are grateful to Dr.R.M.Iyer, Director, Chemical Group for his keen interest 
and encouragement. Acknowledgements are due to Shri T.S.Murthy, Director, Isotope 
Group for kindly providing the data on tritium analysis. Support from Shri S.P.Chaganty 
of Electronics Division for scintillation detector and associated electronics facility, Shri 

J. D.Gupta and Ms Suman Kumari of Computer Division for the 8K PC-based MCA 
system and Dr. M.R.Iyer of Health Physics Division for He 3 well counter and MCS mode 
acquisition set up are gratefully acknowledged. In particular, we wish to acknowledge the 
painstaking efforts and support of the following scientists of the Division in providing the 
reliable data cited in this work, i) Continuous nuclear data acquisition — Smt. G-Leela and 
Kum. Anna John, Ss, Sanjeevkumar, Sunil Jai Kumar, Rakesh Verma, D-Shreevalsan Nair 
and S.C.Hodawadekar. ii) Nuclear and electrochemical instrumentation — Ss. 

K. C.Thomas,J.R.Kale, R.G.Dalavi and A.W.Sahani. iii) Calorimetric measurements — Shri 


A 5 


1 M.Fleisehmann el at, J.Eleetroanai. Chem. Interfacial Electron hem. 261 , 301 (1989). 

2 T.P.Radhakrishnan, Froc. Interdisciplinary Meeting on Hydrogen in Metals, 

DAE Bombay , 1980, p.148. 

3 L. Pauling, Nature 339, 105 (1989). 

4 J O.M Bockris, N. Packham, 0 Velev, G Lin, M.Szklarzcyck and R. K&inthla, Proc. 

Workshop on Cold Fusion Phenomena, Santa Fe, Los Alamos, 1989 ..CE." 

5 J S-Cohen and J.D Davies, Nature 338, 705 (1989)- 



1.00 > 





3 A3' v 



















T P- Radhakrishnan*, R.Sundaresan*, S.Gangadharan*, 

B.K. Sent, T.S. Mur thy t 

’Analytical Chemistry Division 
tlsotope Division 

: "'‘T, 


In continuation of the earlier R&D work carried out in connection with the 
investigations for electrochemically induced fusion of deuterons using palladium cathode 
and platinum anode, a series of experiments was carried out. The following is a summarv of 
results and observations for two such experiments : V' 


PDC— II Experiment 

Start of the expt. 10.7.89 - Conclusion of the expfc. 25.7.89. 
a) Cathode — Palladium cylinder 





Thickness — 0.45 mm 
Surface area - 6.37 cm 2 
Volume — 0.143 cc 
Weight — 1.7 g 

# i %. 


b) Brief history of the palladium cathode 

~ cylindrical type of palladium cathode was obtained and it appears that this palladium 
cylinder had not been used earlier for any tritium work. However, it was used in the 
analysis of the hydrogen gas samples and hydrogen purification work. — Palladium cylinder 
of 1 cm length was used. 

— The palladium cathode was spot welded to a platinum rod 

— The cathode was cleaned by degreasing, using solvents like acetone and was subjected to 
heat treatment at 300°C under vacuum for 2 hours. 

— The anode was platinum gauze of area 25 cm 2 . 

Cell and Accessories 

The cell of about 100ml capacity was made of quartz. The pyrex glass lid was 
provided with cones for introducing thermometer, cathode, entry port for sparging 
solutions with pure argon gas and vent for liberated gases namely deuterium, 
oxygen and DjO vapour for quantitative collection. 

— The DgO vapour collection ; The evaporated heavy water from the cell was refluxed back 
into the ceil with a water condenser and the vapour that escaped from the condenser was 
collected in a moisture trap which was kept cooled with ice. 

- The argon, deuterium ‘ and oxygen were passed through a column of catalyst, which 
recombined deuterium and oxygen back to D 2 O. 

Note: In this experiment, the gases that are emerging after the recombination step were 

passed through a water trap without further provision for recombination over copper oxide 


A 6 

— For temperature measurement, a calibrated thermometer was used. 

— Electro— chemical measur ement : The gal van o— static power source with provisions for 
current pulsing was used. Terminal voltage was measured by a digital panel meter. 

- Materials used in the experiment : 

- Heavy Water : Nuclear grade heavy water was used in the cell for experiments. The 
isotopic purity was 99.87% with an initial tritium content of 170 dpm/ml. (0.076 x 10' 3 

- Electrolyte Electrolyte was prepared 0.1 normal with respect fcoJihOt) by using E. Merck 

grade lithium metal, and 60 ml was used for experiment. -C c 

- IOLAR-2 grade argon was used throughout the experiment aXcarrier gas. 

- The cell was cooled by an externally located fan. 


Tritium Measurement 

— Tritium counting was carried out using LKB system Model 1215, RACKRETA 
system. # </"V 


Samples {2 ml) were withdrawn from the cell during the experiment from time to time to 


tritium content. 

— Counting procedure : Aliquots from 'the above samples have been diawn and added to 
‘ Instage 1 Cocktail 5 with a tofoP©lime of 10 ml for counting the tritium. Known amounts 
of tntiated water from pBS%tandard cells were taken into 10 ml of Instagel as reference. 
The blank sample for tritium was also prepared in Instagel with tritium free water 
obtained through ground water. Along with these background samples, counting was also 
done simultaneously samples drawn from the double distilled water used for diluting 
and making up the samples for counting. Where interesting results were obtained, the full 
tritium spectrum was obtained and compared with the NRS standard tritium samples; 
when the samples gave significant activity, known amounts of the samples were distilled 
and the water distilled was taken for counting. 

— During the electrolysis the cell level was maintained at a constant volume of 60 ml by 
periodically adding pure DjO as mentioned above. Also, the alkalinity was maintained at 
0.1 N by addition of LiOD from time to time. 


The following electrolysis procedure was adopted for the experiment. In the 
beginning, the electrolysis was continued at a constant current (40 hrs). Later the 
electrolysis was conducted using pulsing technique of varying current. It may be noted that 
the pulsing was carried out in this particular experiment only during the day time and in 
the night time the electrolysis was at a constant current. The details of the electrolysis 
parameters such as the values of constant current, pulsing current, duration and also the 
tritium values obtained during the experiment are given below: 

From 12.7.89 to 25.7.89 the electrolysis was conducted in the nights with constant 
current of 1 amp and pulsing of the current was conducted during the day time. 


Degassing Procedure 

Following procedure was used to heat the electrode to recover the deuterium and 


At the termination of the electrolysis the cathode was transferred to a quartz 
heating tube and was connected to a vacuum assembly. The moisture was removed from 
the electrode surface by heating to 80 - 110° C and condensed in a cold trap. After 
condensing the moisture, oxygen and nitrogen gases were recove rtd^j^UHandtod getters. 
The electrode was gradually further heated and the gas released and absorbed 

in a pyrophoric uranium trap. The electrode was maintained at 550 s C for 2 hours till no 
further release of gas took place. Deuterium gas. was released from this pyrophoric uranium 
trap, diluted with about 2.5 l of hydrogen and was conver^rmto water by circulating 
through heated CuO bed This water was taken for tritiunt.qSUntmg. 

Observations : Table IB 

/*>. <ss/ 

- On 11.7 89 between 11 55 and 12.20 hrs. there was£; power breakdown for 25 minutes. 

- The first sample, PDC-II.l, was collecidd^er ‘75 for tritium assay. It gave 1.59 

pCi/ml of DjO. The total volume of DsSrayftexpentneafc at. this time of sampling was: 60 
ml. The tritium activity accounted would bl H 95 4 /iCi, It may be mentioned that the 'total 
DaO (cell volume of 60 ml and make up volume of 40 ml) works out to be 100 ml. The 
excess tritium that has been produced at the end of 75 electrolysis works out to 
1.25 x 10 4 times at this stage. J 

- Second sample was withdraws at 235.65, arid the assay of the sample showed 
that the ac tivity per mb of t he . cell came down to 0.76 pOi/ml. With a view to examine the 
behaviour of the cell, the experiment was continued upta 433 amp. hrs. and intermittent 
samples were taken 

PDC-II.3 (273 3 amp hr.), PDC-II.4(381,5 ), PDG-I1.5(433.32, 

- On 26.7.89 at about 1015 hrs. the lid of the electrolytic cell was thrown out along with 
the anode- and thermometer. The experiment was stopped and a new lid was put with the 
cathode insiae the cell Oxygen and argon mixture was bubbled through the cell and a final 
sample PDC-IL5 was taken on 28.7.89 and was assayed for tritium activity. (0.31 pCi/ml) 

Observations ; Table 10 

-The total activity recovered is 55.52 ftCi during the electrolysis and after the stoppage of 
electrolysis 0 .71 pOi was recovered from the vapour condensate and deoxo- recombination 
trap. Electrode degassing yielded 0.03 pCi and m all 56.27 pCi were recovered. 

- Total input of tritium activity from (60 ml cell volume + 196 ml make-up volume) 256 
ml is 0.02 pCi. 

- In ail out of 256 ml of D 2 0 used, 132 ml of D 2 0 was recovered and accounted for at the 
end of the experiment. 

In conclusion, it may be pointed out that at the end of 75 amp- hrs the electrolysis 
generated about 95,4 //Ci excess tritium (12500 times the input). However, at. the end of 
the electrolysis, that is, after 433.32 amp.hrs. the total excess tritium recovered works out 
to be 56.25 fsC i. 


PDC-III Experiment 

Start of the Expt.6.9.89. Conclusion of Expt.14.9.89 

a) Cathode: Same palladium cylinder used in PDC— II experiment was used in 

b) Cathode preparation: 

To remove deuterium and tritium if any in the cathode the following adopted ; 

— The cathode was placed in the quartz assembly connected to the high— vacuum line and 
slowly heated. Subsequently the electrode was heated under continuous Vacuum from room 
temperature to 850°C. The electrode was kept at this temperature for -two hours. When the 
electrode temperature exceeded 750 Q C, a black deposit was formed" on the cooler parts of 
the quartz assembly. In view of this, heating was continued^atg^O^C for further two hours. 
No further deposit was released at this temperature. Thethl&ck coating formed is being 
subjected to investigations to identify the nature of the^poat. 

— The electrode was brought to room temper&turaybid the vacuum line was filled with 
deuterium gas at 1 cm of Hg pressure. The eiectfode^waa slowly brought to 800°C and kept 
for 3 hours at this temperature. After this dept%itim reduction process, the deuterium gas 
was removed and heating was continued for additional 3 hours under continuous vacuum. 
After this operation the electrode was checked for any further release of gases. No further 
degassing was observed. 

- It may be pointed out that t'he^^cj:,d#posit initially released from the electrode turned 
to a silvery white deposit at thisjt^geiand this coating is under investigation. 

— The electrode after this treatment became shining silvery white and was used for this 

experiment. +; " j? %, 

c) Anode The anode was a similar platinum gauze of area 25 cm 2 . 


Cell and Accessories 

Thedcelht’and accessories remained mostly similar as in PDC— II except for the 

following : ^V 

— Palladium catalyst of standard design was used for recombination. 

— An additional recombination stage consisting of copper oxide column was used to convert 
oxygen depleted deuterium gas into DjO 

— Electro— chemical measurement - The details are similar as in PDC— II experiment. 

Materials used in the Experiment 

Heavy water ; Similar grade of heavy water m was used in the previous cell was used here. 
The heavy water cell volume in this experiment was 80 mi 

- Electrolyte : Electrolyte was similar as in PDC— II experiment. 

- The other conditions such as cooling the cell and tritium measurements were identical 
with PDC— II experiment. 

- Sampling : The procedure followed and the volume of the samples withdrawn from the 
cell during the experiment remain the same. 

In ail during this experiment, 12 samples were taken from the. start of the experiment to 

A 6 

the end of the experiment. Other samples for tritium counting were collected from various 
stages of recombination from the catalyst, the water-bubbler solution and the degassed 


Electrolysis procedure followed was similar to PDO-II experiment. However, 
pulsing was introduced after 6? amp.hrs. of electrolysis at constant current mode. It may 
also be pointed out that the pulsing was done continuously during the day and night for the 
entire experiment. ^y* 

* .™ '%v ^ 

“ On 13.9.89 the electrolysis was reverted back to the constant current mode at 2 amp. for 
3 hours, followed by pulsing till end of the experiment. The pulsing during the period was 
between 2-4 amps. 

— Details of the electrolysis parameters are given in Table 11 
Observations : Table IIB 

— On 8.9.89 between 1300 hrs. to 1350 hrs,,jifch^|ls.was a power breakdown. 

— 9 samples of the cell solution were collected from 7.9.89 to 13,9,89 for monitoring the 
level of tritium activity. The tritium content in the cell solution did not show an increase 
between 7.9.89 to 13.9.89, except that there was a marginal decrease in the samples taken 
on 13.9.89 compared to the sampjeC taken on 12.9.89. In view of this experiment was 
continued with a higher pulsing parameter as that of Fleischmann experimental conditions, 

— During this operation, conspS|| torrent for 3 hours and pulsing for 2 hours were 

— On 14-9.89 around 10QO hrs there was a change in the cell behaviour The bottom of the 
cell was shattered due |o explosion. In all, the total volume of DaO used in this experiment 
was 185 ml which included 105 ml of make-up D 2 O and 80 ml of cell DjO 

— Before the explosion it was noticed that the cell temperature shot up from 7 PC to 80°C 
and the bubbling rate was low. 

Observations : Table IIC 

Due fb the explosion at the end of the experiment ail the electrolyte solution of 
about 80 ml was completely lost and no sample could be collected for tritium assay. Broken 
pieces of the cell were collected. The upper part of the cell and the broken pieces were 
washed .and the sample was collected for counting. The DjO condensate and the 
recombined water from the palladium catalyst, copper oxide and the bubbler solutions were 
taken for tritium counting. 

The total tritium activity counted from all the samples accounted for 2,08,531 dpm 
while the total input tritium activity in 185 ml of D 2 O used in this experiment comes out 
to be 30,170. It may be noted that the activity in the ceil solution could not be accounted 
as it was lost. Even this tritium activity of 2,08,531 dpm. accounted from the recovery and 
recombination of gases liberated from the cell, has shown excess tritium 5.79 times. 

In view of the repeated nature of the explosion at the terminal stages of electrolysis 
it was decided to subject the palladium cathode for a metallogr&phic examination, 
Metaliographic examination was done at the Physical Metallurgy Division by means of 
polarised light microscopy after embedding and mounting the electrode with an acrylic 


tween 13.9.89 
-II experiment 

polymer resin. The palladium cathode was then cleaned with acetone and tetrahydrofurane 
to remove the polymer coating. The electrode on degassing gave about 1.2 ml of gas which 
accounted for 8,580 dpm of tritium .activity. It may be mentioned here that if the electrode 
had been degassed before the metallographic examination the tritium content would have 
been significantly higher. 

— It may be pointed out that had the ceil solution not been lost, significant amount of 
tritium could have been recovered. An important observation from the counting of the cell 
samples was that large tritium activity was not produced in the cell until 13.9.89 
{PDC— III— 12) even at 218 amp. lira. Excess tritium was produced only' between 13.9.89 
and 14.9.89. (between 218 amp.hrs and 286.2 amp.hrs.) whereas in the PDC— II experiment 
excess activity was generated at about 75 amp.hrs. in 3 days of electfoljSis. 

Results of the polarised light microscopy of the electrode 

— Metallographic examination of the palladium cathode used in the experiment and which 
experienced an explosion showed an extensive twinmng within the palladium grains with 
worm— like microstructure. This is suggestive of an, intensive shock— wave impact on the 
metal. Microphotographs of the two samples are attached for reference(Figs. 1—3) 

A similar palladium cylindrical cathode waS also metallographically examined. This 
electrode was given for comparison and did not see any explosion during electrolysis. 
Results of metallography revealed that, it;, did not show any significant twinning and 
worm— like microstructure. 


Acknowledgements are due to Dr. P.K. Iyengar, Director, BARC for his keen 
interest in this work. Acknowledgements are also due to Dr. G.E. Prasad and Dr. M.K. 
Asundi, Head, Physical Metallurgy Division for carrying out the metallographic 
investigations. c-J: 




& ^ A 

Yoi. Of DgO/O I LiOD Electrolyte as 60 ml 

Ttitm£i activity m blank D 2 O/L 1 OD = 170 dpm/ml of D 2 0 

of D 2 0 

=0.076 x 10" 3 /iCi/mi 


N 'w. t' 




Vol. of 

D 2 0 added 



fiCi/ml DjO 









1.25 x 1G< 






3.5 x HP 






2.56 x 10 3 


PDC— II— 4 




1.31 x 10 3 


No sample 





No sample 





PDC— II— 5 




0.95 x 10 3 


0.95 x 103 


source , ^ 

4 .. % 





End electrolysis D 2 O cell-sample 



recovered (PDO— II— 5) v 

V apour— condensat erecovered 



Deoxo— recombmed D 2 O recovered 

16 4* 52 


after of electrolysis 

Vapour— condensate— II recovered 



Deoxo— recombined D 2 O— II recovered 



Bubbler (H 2 0) 



Electrode gas control extracts 



after the electrolysis 

Samples drawn during electrolysis 

Total activity 










A 6 



Amp.Hrs : 

1. Constant current mode 6. 9. 87 — 9.9.89 71.75 

2. Pulsing current mode 9.9.89 — 14.9.89 ' > ; ,, S14.87 

V 486.62 

— -rk - 


Vol. of the electrolyte D 2 0/Q.1M LiOD ~ = 

Tritium activity in blank D 2 0 ■?%, ’ = 

Tritium activity in blank D 2 0/LiOD - = 


— ^ 

80 ml 

166 * 4 dpm/ml of DjO 
166 * 4 dpm/mi of D 2 0 
0.075 x 10-3 ^Ci /ml D 2 0 


Sample No^Cl * 

/ % w 
C S w 

■ 4 


Vol. of DsO 





PDC— III— 4 


195 ±4 






276 ±8 






263 ±8 


PDC— III— 7 



249 ±8 





216 * 4 


PDC— III— 9 



248 * 10' 





256 * 10 


PDC— III— 11 



260 *5 


PDC— III— 12 



250 *5 


no sample 



End of experiment 

* There 

was an explosion and < 
could be taken 

mtire D 2 Q in the 

cell was spille< 

i and lost. In view of this 


A 6 



Initial volume of D 2 O 
Vol. of D 2 O added 

Tritium activity in D 2 O 

Tr i t i urn 
acti vi ty 

= 80 ml 
= 105 ml 

Total 185 ml 

= 166 dpm/ml D 2 O 
= 0.075 x 10-3 /iCi/ml D 2 0 


Jt % 

r,. Sh- 

.... 1 ■ 

(0.0138 pCt) 








i) Cell wash/ broken quartf p|des 



ii) Vapour— condensate recovered 



iii) Palladium catalyst, recombined 



D 2 0 recovered,/ 

iv) D 2 0 recovered- after CuO 



v) Bubbler (HjO) 



vi) Electrode gas content 



extracted after explosion 

vii) Samples drawn during 




viii)End electrolysis DjO cell 




(0.0939 fiGi) 

Excess tritium recovered 

1,77,821 = 5.79 times 

~io tiO 

. = 0.080 /iCi 

•Calculation of total tritium activity does not take into account 80 ml of D 2 0 spilled due 
to the explosion, 


Paper Withdrawn by Authors a 7 


G . V e fikat es w ar an , P.N.Moorthy, K.S.Venkateswarlu, S.N.GuIia, B.Yuvaraju, T.Batfca, 
T.S.Iyengar*, M-S.Panajkax. K.A.Rao and Kama! Kishore 

Chemical Group 
•Health Physics Division 

wr- If 


Recently there have been many reports on the observation hC neutrons, tritium and 
excess heat output from palladium and titanium cathodes electrblytically charged with 
deuterium 1 ' 4 . Steady neutron count rate observed during Operation of the electrolytic 
ceils containing these cathodes and platinum anodes were iij,-sd!tte cases only 1.5 to 3 times 
the background rate 3 ' 4 whereas in some other cases a^mulh higher steady emission rate 
( 10 4 s -1 ) was observed. In some experiments conducted at BARC earlier, neutron bursts of 
two orders of magnitude larger than the background have been reported from electrolytic 
ceils 5 . The possible nuclear reaction responsibly fpr +: the emission of neutrons has been 
inferred to be the ‘cold fusion’ of D atoms*- feasting as D + ) in the metallic lattice of the 
cathode M 

j02 .102 

[ 2 H e 4 ] 31 

2 He 3 + on 1 

We report here a large bupf emi^ion of neutrons (signal/background ratio as high 
as 2000) from a thin ring shaped Pd -cathode during the electrolysis of heavy water at 
relatively low cell currents and: also tritium generation as measured in the electrolyte as 
well as in the water reformed*^® the absorbed gas recovered from the cathode. 

jr . ^ 


The electrpl'Jfig cell design was optimised mamly with respect to observation of 
fusion products- rather than the accurate measurement of excess heat output. This is 
because the quantity of heat required to raise the temperature of 1 gm of D 2 0 even by 1 K 
demands the occurrence of about 10 13 fusions (see Discussion sec.) and hence heat 
measurement appears to be an in— sensitive me.thod of confirming whether cold fusion 
occurs in the cathode or not. Fig. 1 shows the schematic of the experimental set up. The 
electrolytic vessel (vol. ~ 250 cm 3 ) used was made from quartz with a gas tight nylon cap. 
The cap had a number of penetrations for inserting electrode lead wires, thermocouple, 
reference electrode, purge gas inlet and outlet tubes. These tubes were made of ‘Corning 1 
glass. The Pd Cathode used was in the form of a hollow ring 2.5 cm in diameter, 1 cm in 
height and about 0.1 cm in thickness (surface area is 18 cm 2 }. It was pretreated by 
vacuum degassing (<10* 3 mm Hg) at 1073 K for a total period of about 10 hr in three 
heating— cooling cycles (- 3.3 hr/cycle). The bulk density of the cathode was determined to 
be 12 g cm" 3 . The cell configuration was such that the rmg cathode could be charged, with 
deuterium from both sides as it was surrounded from inside and outside by Pt gauze 
anodes. The anodes were loosely sandwiched between pairs of Nation membrane so as not 
to allow oxygen evolved at the anode to diffuse to the cathode surface. The electrolyte was 
D 2 0 of 99.869S2 of isotopic purity containing 0.1 mol dm" 8 LiOD and was kept under 
constant circulation at a flow rate of 10 cm 3 m“ l using a peristaltic pump. The D 2 0 used 
for preparing the initial LiOD solution and subsequent makeups during the electrolysis 
came from a single stock of D 2 0, incorporation of a heat exchanger in the circulation path 
along with a thermostated water bath (273—373 K) served to maintain the temperature of 
the electrolyte (and hence the electrode) at any desired value in this range. Nitrogen gas 


was bubbled through the cel! to reduce the dissolved oxygen level in the electrolyte. A 
saturated calomel electrode along with a lugging probe containing 0.1 mol dm* 3 L 1 OD/D 5 O 
dipped in the electrolyte was used as a reference electrode to monitor the single electrode 
potential of the cathode. Periodic monitoring of the cathode potential was earned out both 
when the working electrode was under polarised condition, and when the polarisation 
voltage was momentarily switched off. 


thermalisation of the neutrons was used fox n detection. This confi^fe:i|tion gave 8 . 6 % 
efficiency for n detection. The output from the detector— preamplifie^agfcive filter amplifier 
tie— up was fed in parallel to a scaler with pre-set time to an oscilloscope and to a personal 
computer operated in a,8K multichannel scaling (MCS) mode with dwell time of 40 seconds 
per channel (Fig,2b). The MCS mode was specifically aLirned at detecting any small burst 

the intense burst release of neutrons the pre-set fcijp#"d5ration was reduced to around 
1000 s to avoid scaler overflow. Any pickup of extranepup signals like those emanating from 
the operation of pump motors, drilling machines,, tesla’ boils, piezo— electric gas lighter and 
fluorescent lamps going off and on was thoroughly checked and the counter was found to 
have good stability and was immune to such electrical disturbances. A constant 
background of about 1.6 counts/s was registered without the electrolytic cell in operation 
for a duration of 10 days before setting up the' electrolytic cell inside the counter well and 
also during times when the electrolysis was continuing but no burst emission of neutrons 
occurred. fV ) 

Accumulation of fcritiujp in the electrolyte was monitored by withdraw! 
periodically 1 cm 3 samples the cel! at about 8 days intervals and measuring 
content using gel liquid scintillation "cocktail” and a Packard counting system. The 
counting efficiency was about 25%. The chemiluminescence effect due to the presence of 
LiOD was seen in the samples and counting was done till such effects died down completely 
and stable count rates were obtained. 

Duringtcirtain periods of electrolysis the electrolytic gases from the cell were 
recombined over a Pt f polyester fabric catalyst and collected in a cold trap, and the 
resulting D20'yvas checked for the presence of tritium. Loss of D 2 O due to electrolysis and 
evaporation was made up by periodic addition of pure D 2 O while losses due to samples 
drawn from the cell for tritium analysis were made up by adding 0.1 mol dm * 3 LiOD 
solution in D 2 0. As stated previously the D 2 0 used for the experiment and for 
replenishments was from a single stock of heavy water. Temperature of the electrolyte was 
monitored by a chromel— alumel thermocouple encased in a glass tube dipping in the 

The cell was run for 32 days mostly at a current density of about 60mA/cm 2 . After 
the completion of the electrolysis run the Pd cathode was disassembled, from, the cell and 
connecting leads and dried. The absorbed gases from the cathode were recovered by heating 
it at 680 K in an evacuated chamber. Prom the pressure volume relationship and gas 
Chromatographic analysis of the deuterium content the volume of absorbed D| gas was 
computed to be 320 cm 3 at STP. This gas was then equilibrated over OuO at 880 K till 

there was no further redtiftmn in vnliimo and 1.1, wafer farmed *■** A 5 <->,,3 \ «tm <-dl!gcted 

there was no further reduction in volume and the water formed < - 0.3 cm 3 ) was coi 
in a thimble cooled to liquid Ng temperature and this water also was counted for T, 

a) Burnt neutron, emission : Fig-2 shows the observed burst neutron emission which lasted, 
essentially for about 40 hr during the operation of the cell for a period of S2 days. This 
figure represents 4000 data points collected in the MGS mode with each of these data 
points being neutron counts in a 40 s time channel. A portion of the steady background 
counts recorded prior to the beginning of the burst emission has been md udC #JjQ show the 
burst intensities in relation to this background. Fig.2a to 2j represent the observed burst 
neutron emission in a ten time expanded lime axis. Upto 14 days^oOlettf olysis (charge 
passed -2.24x10® Coulombs) only background neutron counts were obS^kid and again after 
the 17th day of electrolysis (charge passed ~2,5xl0 6 Coulombs) till tie %nd of the run i,e.„ 
32nd day (total charge passed -3.65x106 Coulombs) only back|rptmd counts were seen. 
During the neutron emission period (15th-l7th day of electrolysis), the bursts were intense 
upto about 14 hr into the emission with intervening quiescent, durations (in which counts 
were close to background) varying from as short as about 120 s and as long as 3 h. In the 
later stages the burst intensity decreased but bursts b|fcjpie'more frequent so much so near 
the end they appeared to be continuous. The maximum burst signal (at about 14th hour 
into the emission) corresponded to 142000 counts jp the 40 s time channel as against 50-70 
counts registered for the background. This corresponds to a maximum signal to background 
ratio of about 2000. Using the 8.6% efficiehcy phthe neutron counter the total number of 
neutrons emitted in these bursts work out, to 18 x 10 s in a period of about 40 hr which 
corresponds to a pseudo average emission rate of 1.3x 10 3 n/s. 

The electrolyte temperature was varied in the initial stages of the run by changing 
the applied voltage to the cell or by changing the set temperature of the thermostatic water 
bath. Ten hours before the burst emission and during the burst emission, temperature of 
the electrolyte was not altered and it remained in between 296 K and 300 K. Electrolyte 
temperature at any time-of electrolysis remained above 293 K. 

b) Tritium in electrolyt.f^and electrode: Tritium counting of samples (Table— I) drawn 
from the 5th— 17th days after the burst neutron emission (sample nos. 4, 5 and 6 in Table I) 
showed an increase from 0.4 to 1.3 3 Low level counting methods (60 minute count 
duration) coupled with a stable system background yielded a standard deviation of only 3% 
on the count rates! The cell electrolyte volume being 250 cm 3 , an increase of 0.4 to 1.3 Bq. 
cm 3 in tritium activity shows an extra input of T to the extent of 100 to 325 Bq., 
amounting tcf 5.6—18.0 X IQ 10 extra tritium atoms accumulating in the electrolyte probably 
as DTG. It is known that at temperatures above 293 K, electrolysis of DjO does not result 
in significant enrichment of tritium in the electrolyte 5 . In the present experiment 
electrolytic gases recombined over a catalyst and counted for tritium have shown tritium 
escaping from the cell. Hence the observed excess of tritium activity points to an extra 
source of tritium, on the conservative side, i.e, not accounting for the tritium that is carried 
away in the gas stream during the electrolysis. 

Degas sing of the Pd cathode at 680 K at the end of the experiment yielded 320 cm 3 
of D -2 gas at STP. By reaction of this gas over CuO turnings at 680 K and quantitatively 
collecting the reformed water and counting for tritium has shown 5.4 Bq. (3x10 s atoms of 
T) .accumulating in the Pd cathode. 

c) _ Electrode potential: Fig.3 indicates the variation of cathode potential and neutron 
emission rate (deduced from the scaler counts obtained with pre-set time) with time. A 
potential of —1.0 V SCE when the polarisation voltage was switched off indicates that the 
electrode in 0.1 mol dm 3 alkali concentration behaves as a near hydrogen (in the present 
case deuterium) electrode. Monitoring potentials some hours into the burst emission period 
revealed fluctuations in the working electrode potentials. The neutron count rate peaks 

A 7 

seem to coincide with the shift of potentials to more positive values by as much as 0.7 volts 
in the cell under "off” and by 3 volts in the "on" conditions. 

d) Heat Output: No discernible extra heating over the background Joule heating was 
observed in this experiment. 


The present study has shown that for identifying whether cold fusion occurs or not, 
monitoring neutrons is a much more sensitive method than monitoring tritium, especially 
when the fusion events generating tritium are not very high .This is because of the small 
value of the decay constant of tritium which requires an accumulation of 10 9 atoms to 
register 1 Bq. of tritium activity. In the present case althou^f%e s ‘fieUtron emitting events 
computed from the integrated neutron counts amounted W'oMyflO® as compared to about 
iO 10 to IQ 11 events computed from tritium data, the neutron, emission occurred with big 
spikes and bursts. As against this, 10 11 events throughTthe tritium channel were inferred 
from an increase of 1.3 Bq/cm 3 extra tritium activity observed at the end of the 
experiment. Since the amount of T held in the Pd cathode at the end of experiment is only 
5.4 Bq, it looks as if the generated T (about 325 Bq) mostly diffuses into the electrolyte 
rather than getting retained in the Pd jggttnxWhich is somewhat akin to the neutron 
emission and suggests the possibility of fusion events occurring close to the surface. From 
the variation in T activity at different times during the course of electrolysis (Table I) it 
appears that T input into the electroly^started after the burst neutron emission occurred. 
This is because the 16th day sample which was taken during the burst neutron emission 
registered the same T activity as the sample taken in the initial stages of run. But during 
the f generation period (17th— 32nd day) there was no neutron emission over the 
background rate, 

Jf . ' 

From the volume of D 2 gas trapped inside the Pd electrode (320 cm 3 at STP 
actually collected), the composition of Palladium— deuteride that would have given rise to 
the burst n emission was inferred to be between PdDo. 3 - 0 . 4 . Using the known diffusion 
coefficients of D < 1>:1 0‘" cm 2 /s it was computed that through the 1 mm thick cathode 
used in this study, D atoms can enter and come out through the other side in about 30 
hours time thus pointing to the attainment of at least PdD 0 -25 configuration in about a 
day’s time. But, as shown by D 2 gas trapped in the cathode at the end of the experiment 
(32 days), only at the end of the 32nd days this composition seems to have been reached. 
The reason for this could be the blackish loose coating forming on the Pd Cathode during 
the course of the experiment which perhaps acted as a diffusion barrier for the entry of D 
atoms. EDXF analysis of this coating revealed it to be platinum which might have arisen 
from anodic dissolution of Pt at the applied cell voltages (which were higher than the 
reversible Pt/Pt, 2 + potential). The reason for not sustaining the burst neutron emission for 
more than three days could be summarised as follows: Those sites in Pd matrix in which 
fusion of _D atoms occurred might have been the really active sites and once D atoms got 
depleted from those sites further replenishment did not take place possibly either because 
of the barrier coating which only favoured molecular D 2 formation from the cathode surface 
with no further entry of D atoms into the matrix . 

A Pd electrode dipped in an alkaline solution and on whose surface D 2 is evolving is 
expected to behave as a normal OD*/D 2 electrode registering a potential of — 1.01 V vs 
SCE. However, in the case of Pd since it absorbs deuterium, the overpotential for Ds 
evolution is expected to increase and the polarisation voltage must be adequate enough for 
evolving D 2 on the electrode surface by electrolysis. However, in the present even after 
charging the Pd for a number of days, the overpotential obtained was negligible though 
degassing the cathode at the end of the experiment has shown that it had absorbed 320 cm 3 


of D 2 gas at STP The shift to more positive potential values by m much as O.T volts m the 
cell coinciding with the neutron count rate peaks means a possible depletion of deuterium 
gas at the surface of the electrode. Thus monitoring the cathode potential has thrown 
some light of what is happening to a Pd electrode which has absorbed deuterium 

in general very few groups have been able to reproduce the iarge heat output 
reported by Pleischmaon et ai ! and Mathew et al‘- Even to match a cooling .f-ate of 0.1 K 
per 'minute in a cell of dimensions reported m the present paper if would req lire the 
occurrence of fusions at a steady rate of about tQ !C /s. As discussed above, the observations 
of n bursts revealed that the .occurrence of fusions is not at a steady- ilyiP of the desired 
magnitude and hence there is no wonder that the excess heat ouipjii wfcat undetected. In 
our laboratory we have carried out several runs on electrolytic £<ff» lifemg Pd cathodes of 
various shapes and dimensions (including ' 

and Pt anodes in 0.1 mol dm 3 LiOD/ DjC 

cylinder and 1 cm long) 

cathodes and Pt anodes 

along with controPald 

m 0.1 dm 3 LiOH/ HjjO purely with a view to observe the exclss heat output and were not 
able to observe the same within the limits of uncertagf^y of our calorimetry, which is 
estimated to be about ±10%. 


The present study has shown edabmai for cold fusion phenomenon in an 
electrolyte cally charged Pd matrix in terms of neutron and tritium as the signatures. The 
tritium channel seems to be favoured ovyh; the neutron channel, During the period of our 
experiment a total of 10 u fusion events leading to tritium generation were observed 
whereas the neutron channel accorfrtt^iu "fdr only 10* fusion events. No significant heat 
output over the Joule heating cnalddae observed The present study lias also revealed that 
the effect is small and not a lUstamcd one Energy production from cold fusion of 
deuterium in an electrolj^c^^rcnarged Pd matrix in a sustained manner may require 
more systematic exploraii'pn^ to identify the various parameters governing the occurrence of 
the process, not the leas^n^poriairi among which is the proper pretreat, meni of the 

Acknowledgements A? 

We are 'thankful to Dr.R.M Iyer, Director. Chemical Group mid Dr.J P Mittal. 
Head, Chemistry Division for their encouragement and support for this work, to 
Shri.S.D.Soman, Director, Health and Safety Group for guidance in tritium analysis and to 
Dr.P.R Natarajan, Head. Radiochemistry Division for making available the neutron 
counting setup We are grateful to Drs, R ILlyer and J K. Samuel of R&dtocbemistry 
Division for useful discussions and to Shri V N.Rao of Chemistry Division for providing the 
necessary instrumentation for electrode potential measurements The authors wish to 
thank Shrs Balkans Jayswal of Computer Division for helping with his GraphXY routine 
to plot with optimum resolution the large number of neutron counts data gathered in the 
MCS mode We gratefully appreciate the tireless devotion and patience with which the 
following personnel from Water Chemistry Division and Chemistry Division kept these 
experiments continuously running over the long time period . S/s B.R.Ambekar 
T.P Baian, A D.Belapurkar K B Bhatfc, K-Si.Bhide, A S Gokhale, Dr Hari Mohan, 
DRJoshi, P.S Joshi, Dr K N G Kaimal, S,B.Karw«sr, A.S Kerkar, Or M K umar 
H.S.Mah&i, Mam Max an, N.Masoj, M.O Mathews, R.K Misra, Dr A Sat vansioorthy and 
K N .Shelar 


A 7 

! M Fleischmann, M Hawkins arid S.Pons, J. Electroanal. Chem 261, 301(1989) 

2 S E Jones, E P Paiiner, J B Czirr, D L Decker, G L Jenson, J M Thorne, S F Taylor 
arid J Rafelski, Nature 338, 737 ( 1989 ) 

3 K S V Santhanarn, J Rangarajan. O'Neil Braganza, S K Haram, Limaye and 

K C Manchal, Indian J Tech. 27, 175(1989;. J..-T 

4 C K Mathews, G.Penaswami, K.C Srtnivas, T.GnanasekaracO%®Hajan Babu, 

O.Ramesh and B.Thiyagarajan, Indian J. Tech. 27, 229 (1989)* .7. 

K Iyengar, "Cold Fusion Results in BARG Experiments".. ,PFoc Fifth International 

..r r> ... xi i i t i , . - _ 


on Emerging Nuclear Energy Systems held at 'Marlsrulie, FRG. July 3-6. 

1 I I tlf t l n ' ✓ , 4 * 

1989 Published bv World Scientific. 

e 09891 f* , v i 


tabu; k 

W «j| 


, 4 M 


■ ’■■■■„ 

Duration of .-**>' ■ 

•“*' — 

. Coulombs Passed 

Tritium Activity 


• Electrolysis 

{ Cumulative) 


(d 4 X 

Sf /: 0 

(X 10-«) 








’C^ 16 


2 36 

3 32 





4 94 





*> • 4* 4 


*> K T. 

r, on 

t / 

*J .44 


“ Standard deviation: ± 3% 







3, 4 -CATHODE RING (Pd) 







FIG, 1 





H. Bose*, L.H. Prabhn*, S. Sankaranarayanan' 

N. V eeraraghavan*, P.V. Joshit, T.$. MurthykBK 

K-G.B, Sharmat 

’Reactor Operations and Maintenance Group. 
I Isotope Group. 




In continuation of the R&D work carried out earlier on .efectjbchemically induced 
fusion of deuterons, a series of experiments were carried outaismg palladium cathode and 
platinum wire gauze anode. Some of the mam features of the experiment were as follows; 

- The electrolysis experiments were conducted us^|0.1 M LiOD solution in DjjO 
using Ixlxl cm cube of palladium as cathode and platinum wire gauze as anode, 

- The experiment lasted for seven weeks during which data on heat measurements, 
tritium production and neutron emission were collected. 

~ Ihe amount of deuterium— loading on palladium electrode was measured at the end 
of the experiment. ^ 

- Blank experiments with stainless steel cathode in place of palladium and H 2 0 in 
place of D 2 0 were also carried opt. v\ 

Experimental V 


Cathode f 

(a) Palladium Cube Cathode; Thfe spectrographic analysis showed the palladium content to 
be 9S.2%, The impuriti£%#ejpe' mainly Cu, Ca, Fe, Pb, Ni, etc. The cathode was heated 
under vacuum and cleaned eiectrochernically before use. 

(b) Stainless Steel Cube Cathode: 1 cm x 1 cm x 1 cm cube of stainless steel (Type 3004 L - 

Fe 72%, Or 18%, was used after polishing the surface and cleaning with distilled 


The palladium cathodes after use in the electrolysis experiments were generally 
degassed at high temperatures (850*^0) and heated in vacuum for a few hours for reuse. 


The electrolyte solution of 0.1 M LiOD in D 2 0 was prepared by dissolving lithium 
metal in D 2 0 having an isotopic purity of above 99. 86% w/w. 

- The volume of solution used in the electrolysis cell was 140 ml 

- in the case of some of the blank experiments with H jO, the electrolyte solution of 
G.1.M LaOH in H 2 0 was prepared by dissolving lithium metal m pure double 
distilled water. 

Experimental Assembly 

Pig.l gives a schematic diagram of the assembly. Fig. 2 gives the details of the 
assembly used along, with Fig. 1 for recovery and conversion of gases (D 2 and O 2 } liberated 
during the _ electrolysis... It may be mentioned that a palladium catalyst was used for 
recombination of D 2 and 0 2 and hot copper oxide was used for conversion of Dj into D 2 G 
- The recovery assembly enabled assessment of recombination of D 2 and 0 2 emerging 
from, the electrolytic cell and also permitted estimation of the absorbed D 2 gas 


A 8 

liberated from, the palladium cathode .after the electrolysis was terminated 

;rimental assembly in Fig. 1 (Cell) was covered with thermocole insulations 

The expe; 

open to ambient air during the initial stages when data on heat measurements were 
collected. Subsequently the cell was cooled only by air and during the later part of 

metrical heating 

the electrolysis, the cell was cooled by cold water. 

A resistance wire heater was placed in the cell to collect data on 
without electrolysis. 

Temperature of the electrolyte solution was monitored by a Pt-100 RTD placed in 
the cell and it was recorded continuously on a chart. In the initial stages of heat 
measurements, a mechanical stirrer was used for imping, the electrolyte. 
Subsequently the cell Was stirred with the bubbling of nitrogen gas from a cylinder. 


Jf % 


Electrolysis was carried out at constant current and variable voltage. At the end of 
1365 ampere-hours, current was changed to pulsing, mode (the cycle being G.5A for 1 
second followed by 2.5A for 11.5 second) for about three hours and then to the cycle Q.5A 
for 1 second followed by 4.5A for 1.5 second. g /■'. 

— The latter lasted for about ten minutes when an explosion occurred inside the cell 
resulting in the cell lid being thrown out with the electrode assembly. The 
electrodes were intact but were fouh4)'electrically shorting. The electrolysis was 
terminated and the I >2 gas released by the palladium cathode electrode was passed 
through the assembly of palladium catalyst and copper oxide for conversion of D 2 to 

Neutron Detection 


— For neutron detection, a BF 3 neutron detector with a counting efficiency of 5 X 

10’S cps/ fast was placed below the electrolysis cell. An additional BF$ 

detector was kept about two meters away from the cell to monitor background 
neutrons. - 

Sampling and Assay 

— The cell electrolyte as well as the D^O collected in the various cold traps in the 
attested assembly were periodically sampled during the course of the experiment, 
for tritium measurement. The D 2 gas released by the palladium electrode (after 
conversion of D 2 O) was also analysed for tritium content. Ail the tritium 
measurements were done using LKB system Model 1215, RACK BETA-41 system. 

Blank Experiments 

— Blank experiments were earned out following a similar procedure using a 
combination of palladium and stainless steel cathodes. DgO and II 2 0 electrolyte 
solutions, and following the electrochemical parameters m given in Table I. The 
approximate duration of each blank experiment was about 16 to 20 hours. 

Heat measurements 

For measurement of the heating rate in the electrolysis cells, the cell temperature 
rise _ versus time data were collected for different current density values till the cells 
attained thermal equilibrium. Also steady state cell temperature data at different stages of 
electrolysis with increasing deuterium loading in palladium were collected. 


A 8 

Fig. 3 gives the variation of equilibrated cell temperature rise with Joule heating of 
the cell. The cell thermal sensitivity expressed as the ratio of steady state cell temperature 
rise (with respect to heat sink temperature) to Joule heat input power, remained almost 
constant, in ail experiments with palladium/ stainless steel as cathodes and D 2 O/ H 2 0 
solutions of lithium as electrolyte. This seems to indicate that the electrolysis cells 
behaved similarly so far as their heat production characteristics are concerned." However 
this observation does not specifically resolve presence or absence of excess heating in the 
cells To examine this aspect further, the cell temperature versus time data were analysed 
using a non— linear least square fit algorithm to obtain the cell calorimetric heating rate 
employing a "time— constant" method. The calculation method ^Sriverified to yield 
satisfactory results (within "about 10%) for simulated electrical - resistance heating 
measurements of a typical ceil configuration. Results showed Aa|^c'om.pared to the cell 
Joule heating, the cell calorimetric heatings were consistently, higlver within experimental 
and cairulalional errors for all the cells as shown in Fig. 3. 

Table I summarizes the results. The ratio between lie calculated cell calorimetric 
heating to the Joule heating rates varied in the range l A to 2.3 for Pd-D 2 Q system, 1.1 to 
1.6 for SS-D 2 O systems, 1.6 to 1.9 for Pd-H 2 0 sysiems 1.5 to 1.9 for SS-H 2 O systems. (It 
was also experimentally established that tl|g|( feg"' and 0 2 gases do not recombine 
significantly within the cell volume which coulRTead to cell calorimetric heating rate higher 
than the Joule heating rate). 

■ ■ 

lor the Pd— D 2 0 ceil, the excessj,over Joule heating rates computed were used to 
estimate the extent of known nuclear., fusion reaction rates. To allow for possible heating 
contributions from any chemical reactions in the Pd— D 2 0 cell, the excess over Joule 
heating rates computed for the blank experiments with Pd-H 2 0, SS— D 2 0 and SS-H 2 0 cell 
were least square fitted as a straight line function of cell current. Due to similarity in cell 
geometry and cell contends JPIk ilast square ft results were assumed to be applicable for 
the Pd-D 2 0 cell as well: - 

If one were to%assifine that the extra excess heating over Joule heating rates thus 
obtained for the Pd-AKPO cells for various cell currents was due to known nuclear fusion 
reactions with deitiefbns, then one would obtain the fusion rates per second per d— d pair to 
be approximately. 7 3 X 10“ n assuming only D(d,n)He 3 reaction, 6.2 X 10" 11 for only 
D(d,p)T reaction and 1.0 X 10 _,J for only reaction. 

When the Pd— D 2 0 main experiment was conducted in current pulsing mode with 
low heat, sink temperature environment, the cell experienced a temperature transient of 
about. 25*C over a period of 8 minutes. The cell had a mild explosion which dismantled the 
cell configuration. 

Subsequently blank experiments with near simulated heat transfer conditions (as in 
the main experiments prior to the temperature transient) with low heat sink temperatures 

were done. 

These experiments showed that typically for an electrical power input of about 40 


(a) the cell temperature stabilizes at about 27 C C in about 15 minutes with a low heat 

sink (water bath) temperature of 4°C, 

(b) when the heat sink water was drained (as happened during the explosion event) the 
cell temperature increased at the rate of about 2°C/minute, and 


A 8 

(c) a low temperature heat sink in the form of ordinary ice (as observed prior to 
explosion) was not efficient to reduce the cell temperature substantially. 

Further with an electrical power input of -1 15 watts into the cell (as existed during 
the current pulsing conditions) the cell temperature could rise by about 6.5°C/minute even 
with low (water bath) heat sink temperature of 4— 5°C. 

These experiments thus showed that the observed cell temperature transient could 
perhaps be due to inadequate heat transfer from the cell surfaces or due to high electrical 
energy input into the cell during the pulsing experiments The •reason for the mild 
explosion however has not been understood. % 

Tritium formation j' 

» ••ess? 

For a proper assessment of production of excess tritium in the electrolysis cell, 
known data for the input tritium in the total D 2 O used in the experiment, and for the 
tritium analysed in the various samples including that from the palladium electrode, were 
used. Table II gives tritium results. For the unaccounted D 2 O i.e. D 2 O which could not be 
accounted in the daily D 2 O material balance by make up and recovery, a tritium value as 
shown in Table II was taken, based on the 4rHuitn 'value experimentally obtained for the 
recovered D 2 O. Thus it is seen that a net tritium, excess corresponding to 50 percent of the 
total input tritium was observed. The measured excess tritium production of 0.0348 /tCi 
for the main electrolysis experiment corresponds to an average fusion rate of 2.3 x 10' 1S 
fusions per d-d pair per second assuming D((d,p)T is the only reaction. 

Neutron emission 

if .,-T' •: w" 
i ,i¥" |j 

In the initial stages ,gf ; e|eclrolysis with Pd-D20 cell, 17 neutron bursts lasting for 2 
to 55 minutes each were obsefved. The integrated neutron emission in the bursts varied in 
the range 5.1 X 10 s (burst period = 2 minutes) to 5.4 X 10 s (burst period = 8 minutes). 
The background neitrolK level recorded was about 20 neutrons /second during these 
measurements. However, there was no indication of neutron emission later in spite of the 
increased loading of-palladium with deuterium. In fact, analysis of absorbed deuterium in 
the palladium electrode after the termination of the experiment gave a value of D/Pd > 2.2 
which sho^fd~Ihat the electrode was saturated with deuterium. The measured number of 
neutrons emitted in the bursts observed in the mam experiment indicated an average fusion 
rate of 2.2 X IQ“ 21 fusions per d-d pair per second assuming D(d,n) 3 He is the only reaction. 


(a) It may be noted that the fusion rates calculated for the experiment via known 
nuclear reactions do not appear to tally with each other. 

(b) During seven weeks of electrolysis, the Pd-D^O cell experienced m all three 
explosions two of which lifted the cell lid. 

(c) The amount of D 2 gas released by the palladium electrode after the termination of 
the electrolysis (duration seven weeks) as well as that obtained by heating the 
electrode to 800°C, was used to calculate the deuterium— loading value expressed as 
D/Pd. It was found to be not less than 2.2 The deuterium loading value is an 
important parameter for calculating the fusion rale per d-d pair. Since some of D 2 
absorbed by palladium was lost to atmosphere during the last explosion, the actual 
deuterium— loading value (D/Pd) is probably close to 3. 

Additional experiments are in progress using palladium sheet electrodes to study the 
cold fusion phenomena further. 


A 8 


The authors wish to acknowledge the assistance rendered by Ss. G. Bharadwaj, 
M.D- Dharbe, M P. Kini and Jacob Joseph for the instrumentation and fabrication of the 
experimental set up. The authors also wish to thank their chemist colleagues for 
round-the-clock collection of experimental data. 



Electrolysis Cells With lxlxl cm Cube Cathodes Analysis of Cells Heating Rates 

A 8 

• O 
M t- 
«) M 




5 __ 
■la ■ 

<u . 






-a a 

O 4! 


•4*3* M 

CS 3 


4-3 <1? 







f 'A 



• o 

002 : 


<M W <N W '<N 


•T3 a “i'A ] ' ij ^ 't'l* '' fi 

a *aQ0i 
co gg m m m 

■to j 

in* n- co 

CO CO i 
CO ^ ’ 

f CO 

t 04 CQ ^ to 


W ^ 




OS 05 CC 

cc oq oo 

00 oo ■«$* 

01 co -sr 

CN C-t C4 

co r— oo 


• csi et ea __ 

’ — * | | l ' — * 

R co 

^ CD r**t 

m « 
d *-i 

^ m mm co 

<D CM ^ <£> m 

CO C7> O 
■#*4:. CM' <0 

VO ® 

m o 

V"** :<£SJ *•— « 1 1—4 

r— i CO r- * — « C4 

x so o 

<“jO SD 


Mt 1 

-#• | <- i.o <© to 

t— * CO <X3 


: '" . 

C - 


23 v 

IO QJ o OO Tt 

O oo o 

co vo 

.-Hi d to 

CM cb 

& )§ 

<8jr t® J|F 


tf» 0» Mg :,. 

to CM CM 

O <N 



OO *>^ 1*3 '<&"*%# 

•CM I— 

cb tb 

f"**4 O 


^ O' 0C- CD *— « 

& % % 



CO cri 0> <0 r-H 
CO CO ^ lO ^ 

CM CO tO-' 

CO 10 
^ CO ^ 

| rf O OO CD — 

■’tf* 1*0 I"** 

.^O ,, 



[ m uo i.o cm -X' 
tO SO I s - oo so 

*^4 CN4 

1,0 to oo 














ca oo 

" o 



Ct C 4 

irr*- *t* 
*-w *«M 





m m 

■tO IN* 

o» -■ 


04 LO 
10 OQ 

oq 04 

CO *>» 

oi co 

oq oq 
co l6 

o o 

CO ^ 

■ R R 
oo cm 
uo 3h* 

CD *** 


«S» CSt 


•"O '*“0 

a* au 

<» o> 
oo- oq 

cri oi 

> CM 

£ 21 

<s 3 

,+* 4 3> 

Nr3 ^ 

*”5& L-4 



j- cn 

rU- "2 


B a 

TU iM 
v at 

ti m 
s3 = 


is} S 


i 4 






A 8 

Table n 

Tritium Balance 

Tritium Input 

Tritium Result 


a) Initial volume of D^O in the cell 

b) Volume of DjO added for make up 


Tritium Input from (a) & (b) 

Tritium recovered in the experiment (in pCi) 







Tritium m the electrolysis cell at the end 
Tritium in the vapour condensate 
Tritium from recombined gases 
Tritium from D^G formed m the CuG" 

Tritium from electrode degassing 
Tritium in samples during electrolysis 
Tritium in unaccounted D 2 O (359 ml) 
lost during the experiment (i,g; 1 ?359 x 181 

Total Tritium in (a) to (g) in II 

III Excess tritium (pci) , 















2. Pd. Jem. x J cm. x 1 cm. 

3. PI. WIRE MESH 2 cm, 
4 cm. HEIGHT, 

D 2 0 LEVEL 



- A L. TRAY 


A 9 


-TSMurthy , ’*T.S, Iyengar, B.K Sen* and T.B. Joseph* 

* Isotope Group 
Health Physics Division 


42 #-' 


The report summarises the methodology arid techniques adopted for 
determination of tritium content in various samples obtained during the initial sets of 
experiments conducted at Trombay in connection with studies~:pn-the feasibility of ‘Cold 

1 usion 

The analyses were carried out at the Isotope Division add Health Physics Division. 

Sample Preparation Technique 


h. sS > 

Sample preparation techniques inyblyld.. %se of the appropriate scintillation 
"cocktail 1 and wherever applicable, the sarnpie& were distilled before use. Diluting the 
sample with double distilled water, even though reduced the number of signal pulses per 
unit volume, of sample, helped in reducing the pH as well as quenching impurities present 
in the sample. Alternately the sample codd be kept for ‘chemiluminescence cooling' so that 
the contribution from the same, if any, ."is reduced to a negligible level with respect to the 
sample under counting, - d 

4QK free vials are jse^ffcrtow background. Use of Dioxane as solvent was avoided 
wherever direct counting yiasbadopted, as it tends to show chemiluminescent properties 
when used in certain '%yaiiples. The commercially available scintillation 'cocktail' 
IN STAG EL, (containing a surfactant such as Triton — X— 100) was found to be suitable for 
counting m such situations. 

Standardized procedures involved addition of 0.1 to 2 mi of sample in appropriate 
volume of scintillator in cases where the count rates were high. Larger quantity of sample 
was taken (8 to 12 mi of sample) in the case of low level samples for better detection limit. 

Counting System 

There are several LSS systems available for estimation of low— energy beta emitters 
like 3 H, !4 C etc. In the present case the LSS system manufactured by M/S Packard Inst, 
Co. (Model 4530) and LKB system (Model 1215 RACKBETA— II ) were used (Some 
samples were analysed m an alternate LSS Packard 4530 too, for confirmation of reliability 
of the method.) The LSS used has facilities for automatic quench correction. The probable 
errors due to interference due to chemiluminescence is also avoided by adopting 
appropriate chemical counting methods. 

The system stability is checked everyday using sealed tritium standards and sealed 
background samples in a 'calibration' mode, so that normalization is effected by adjusting 
the two PMT voltages automatically by the LSS system itself. Standardisation and 
efficiency of each sample is determined using quench curves developed for each batch of 
experimental samples by using Quench Standards of the appropriate chemical form. 


In the case of heavy water samples used for each set of electrolytic experiments 
initial samples were drawn and kept aside and counted along with the samples drawn 
during the course of the experiments. 

The rooms where the experiments are conducted were constantly monitored for 
tritium contamination, from air. 

In almost all the experiments samples were drawn at appropriate intervals to follow 
up the trend in tritium concentration values. 

Experiments which have shown definite increase in tritium concentration values are 
listed separately. 

In all the tritium measurements the following factors were considered in arriving at 
the excess tritium content produced (if any) in the experiment. 

a) initial tritium content in the heavy water used for each experiment, 
bi concentration of tritium content due to electrolysis 
c) concentration effects due to make-up volumes of the heavy water. 

> \v* 

Materials And Method 

Tritium Content in Heavy Water Before Electrolysis: Heavy Water used for 
electrolysis experiments at the Analytical Chemistry Division and Reactor Operations 
Division are analysed for tritium ; cphteht. In these two Divisions (ACD and ROD) almost 
ail the experiments were conducted using the heavy water from these stock solutions. 

Tritium Content Mladium Cathodes: For the electrolysis experiments at the 
Analytical Chemistry Division palladium metal has been used mostly as cathode in 
different shapes. Befol%/lhe cathodes were prepared samples of palladium metal was 
collected for tritium assay. In addition, samples of palladium metal and palladium salts 
from Radiation Technology Division were also collected and analysed for tritium content. 

Palladium metal was dissolved in aqua regia, while palladium salts were dissolved in 
water and nitric acid. From these solutions suitable stock solutions were prepared. 

These include 

a) acidic solutions 

b) neutralised solutions 

c) diluted solutions and 

d) diluted and neutralised solutions fmaily made for counting. 

In addition, dummy stock solutions were also made with the reagents used except 

The counting was continued for a period of one week. Acidic solutions have shown 
initially very high chemiluminescence; diluted and neutralised samples after prolonged 
counting have shown that the palladium and palladium salt do not contain, any tritium 

Checking of Lithium Electrolyte Generally, lithium, salts (LiOD) have been used as 
electrolyte (0.1 Molar) in the electrolysis experiments. In view of this, the lithium solutions 
were checked for 

a) tritium contamination, 

b) chemical effects including chemiluminescence etc., 
in the counting of samples from the cells. 

For this purpose lithium deuteroxide solution (02M) were prepared with known 
heavy water (D 2 O) stock solution and analysed for both a) and b). 

From the counting data the following observations are made: 

1 Immediately after preparation, high chemiluminescence was observed. However with 
passage of time this chemiluminescence came down to negligible amounts (in about 10 days 
time). When these lithium stock solutions were neutraiised.the, bhejniluminescence effect 
was greatly reduced to negligible values. "V ' 

2. When these samples were subjected to distillation and water was collected and 
counted, no chemiluminescence was observed. The distilled samples have shown the 
original tritium content of the heavy water (DjO— 4) used for this purpose. Therefore no 
tritium contamination was noticed in lithium salts used for the electrolysis. 

3 However, it has been observed that tha^ectrolyte samples which have been subjected 
to few days of electrolysis have shown small amounts of chemiluminescence when compared 
to unelectrolysed lithium deuteroxide ^ampffes. Initial chemiluminescence which was 
observed in. some of the electrolysed samples was found to decay rapidly within 24 hours. 

4, Therefore, tritium, content in tAr final samples from cells was confirmed by counting 
the EhO distilled from these samples 

Quench Corrections /A * 

Due to the presence of chemical impurities even in trace quantities chemical 
quenching possibiliri^ fkist, which in turn reduce the pulse output. In impurity quench the 
components which A Seh the sample do so either by competing with the fluors for energy 
transfer, or by chemically interacting with the fluor molecules to make them less reactive 
to energy transfer Oxidizing agents at high pH can alter oxygen atoms in many of the 
fluors so that the fluorescent properties axe also changed. 

In colour quench, the quenching component absorbs the photons produced by the 
scintillation process, before they can be detected by the PMTs. Colour quench usually does 
not interfere with the scintillation process but exerts its effect by preventing the photon 

being seen by the detector system of the LSS. 

In both the above cases the resultant effect will be a compression of the beta 
spectrum, of tritium and reduction m pulse output, thereby reducing the efficiency of the 
LSS. Thus it becomes necessary to correct for such quench errors mi arrive at the 
efficiency of each one of the samples being counted. 

Quench corrections axe carried out by any one of the following methods 

i) Standard Addition Technique wherein a known amount of standard of appropriate 
concentration is added to the sample which determines the efficiency for that particular 
sample. In this ‘spike method 1 the sample becomes irretrievable. 

ii) Sample Channels Ratio Method effectively takes the ratio of two preset channels of 
the beta spectrum for the sample and comparing it with a ‘calibration curve* developed 


A 9 

with a series of known quenchers of appropriate chemical form whose efficiency have been 
determined previously. For very highly quenched samples this technique is not quite 
suitable as the ratio of the two channels will not reflect true picture of the actual, 

shooting a gamma emitting pellet which occupies a place very near the counting vial for a 
very short period and produces compton electrons which also undergoes similar quench 
effects as that of the sample. Here again a ' calibration curve ' with a known set of 
‘Quenched Standards 1 of appropriate chemical composition helps m determining the 
efficiency of the sample being counted. /**’ ^ 

A later version of the same technique developed by Donald Horrocks uses the 
compton edge count rates for fixing the channel position for the quenched standards and 
calls it * H Number*. Each sample being counted will; record its H number which provides a 
method for determining the efficiency. 1 /%# 


J \ % 

Table— I gives the tritium contend ift- the heavy water stock solutions used for 
electrolysis experiments conducted at Analytical Chemistry Division and Reactor 
Operations Division, and Heavy Watfh Division, BARC. 

Tabie-II gives the data" l^ftaining to 10 electrolysis experiments conducted at 
Analytical Chemistry Dmsicp^md Reactor Operations Division, BARC. 

Table-Ill Froip>th4sis of values given in Table II the data where excess tritium has 
been observed in the ij|dtrpiy te solution after electrolysis have been summarised and given. 

Table-IV The data from experiments conducted at Heavy Water Division, BARO 
are tabulated. It may be pointed out that the correction factors due to make-up volume 
contribution in the final tritium content have not been reflected in the table. 

Table-V Information regarding the experiments conducted at various Division/ 
Sections have been given. The samples counted from these experiments have not shown any 
apparent increase in tritium values. 






DO - 1 


DO - 2 

DO - 3 


V - 4 

DO - 5 

DO - 6 


DO - 7 

D 2 ° - 8 

{in uCi/ml) 

1.16 x 10 

0.4S9X 10 

0.845x 10 

# \ | 



0.07 6x 10, 

0.117X 10 


0.27 X 10 

0.045x 10 


0.055X 10 


This values was given by HWD as the average value 
of initial^fcitium content in D^O used for the 
experiments HWD (B) and HWD (C) . 




Tram ccurajJGMT* a Haumca, aemsm wvisich otxrdsots 

XU eaUwdes: palladium metal XU anodes : platoura 

Inltml 1)0 
yslms aided 
at BJi tmi 


Final Tap. tes. Beetta 
0,0 lytfi. 



• _ . 

D^O used Ck-rtrc- DO acti Set 

uCi/al lysis end nty after trittma 

A sassple velum actrtifjc*' js 

uCi/kl correction 

B oCi/el 

fXjvUil 45 

ftptiitt 65 

BtpMiiil 60 

fi«etULU« is 


%kr »• 




0,1 H 


0.045x10 0.45 l.Tfc^V 


„» ’%_# 'St* _} 

0.48MS *.J» Tp.SSSxlO 

..... 'll 

g&p ” 



Excess tritium 





0.1 « 



Essmm txikim 




0.1 ft 

NaO D 

0. 076x10 f il.ixK _ 0.I8M0 

Steess tfitim 
pot oJssenred. 




0.1 8 


O.Uixlo’ 3 0431x10* 3 





boiled oil. 






a&»3 H 


0.114x10 3 

. . . -3 

O.€$te0 - Black coatiag 


fanned co 


. m, 

%&# § 






t# 1 


0.1 H 

0.076x10 ' 

o.imo ” 3 


0.922x10 - Stopped due to 

K-» ? 


/ 5 > 

Jr s> '%. 

40 \j/g 61-83 


0.1 H 

0.076x10‘ 3 

0.09M0* 3 



0 406x10 - Excess tritiin 


S|*v \ 


001 observed. 

RC-III '* 




0.1 H 

0.076X10’ 3 

0. 239X1 (f 3 

0.426X10 3 - Experiment 










0.1 M 

0.076xlo’ 3 

2 , 6 xl 0' 3 

-3 -3 

0.1X10 2.5X10 Excess tritium 








0.1 M 

O.O76H0* 3 

6.83ao* 3 

0.098x10 3 6.73x10 3 Excess tritiua 





Ml E - III 


s® - sashes 

./sF •«> 






0.1 H (?.O76xl0’’ 3 

LiOH ---- 

2.6xlO~ 3” 3 2,sao”i 

J&wts* tritiuo 






Jg _ * 

yyl 0.076X10 

6.83X10' 3 

O.OOSxlO' 3 6.73x10’ 

1 Exse3s tritim 




Division No. of 

Expts . 

Physic. Chem. Sec. 2 

No. of samples 


a ft 

Pd_-Hi system: 
Lip© 10.1 M> : 
c membrane 
electrodes : 
Vol. and cell 

dimensions changed 
in different 
cells . 

Pt-Pd system: 

LiOD 0.01 M 

Control Exp. 

with B SO 
2 4 

Pt -Pd system: 

LiOD 0.1M 
Electrode size 
varied : 

One cell exploded. 





LSS Used 

System Bkg for full Channel 
(0-19 keV} 

Real Time Monitoring for 
ambient gamma fields? 

System Stability Check ? 
Corrections made for 

Packard {Model 4530) 

<12 cpm 


Yes { in 3 rd channel } 


* <av 

Yes {ever y cfasy 1*S 

a) Chema-. -and'-- Photo- 

b) H 2 0|^0 2 

presence if any 

cjJs^sBical and Photon 

dfr Volume Quenching {DO) 


■T ; L. 

VW Spectral Index Sample 

Quench Corrections 

(SIS) and Spectral ExtB^nal Standard f SES) in the 
Packard LS System. Crd^s check with Sample Channels Ratio 
(0-19 keV and 2-19 keV regions) is also done. 


A * 

Ref , 

No. of Samples 

Heavy Water' D i v . * 




Chemistry Div. * 







H !~ 

V t t 





H/HN Series 


t 1 




#• f 




i r 




, f 




Analyt. Chem Div, 




Metallurgy Div. 




* The identification and interpretation of various 
sets of experiments are given by the respective groups 
separately, elsewhere. 

NOTE Chemistry Division Experiments are planned such 
that different parameters are studied, independently by 
different groups. 


A 10 

S.K.Malhotra, M.S.Krishnan and H.K.Sadhukhan 
Heavy Water Division 


Electrolysis of heavy water has acquired great importance during the last few 
months m view of the discovery of the phenomenon of cold fusion 1 * 2 . Though the initial 
papers reported measurements only of neutron emission and/ or heat, lately tritium also 
has been reported to be forming via this phenomenon 3 . However any tritium measurements 
carried out should be corrected for the isotopic enrichment of tritium during electrolysis 
since it is a well known technique for hydrogen isotope separation. In this short note we 
have attempted to give a complete material balance oLtritldm escaping the system m the 
form of DT gas and also as DTO vapour. Tritium produced in excess of what is predicted 
from this equation may be attributed to nuclear fusiqjileactions. 


Basic Material Balance Equation » ^4T\* 

Consider an eiectrolyser (Fig.l), in Which the deuterium gas formed is being let out, 
D 2 0 vapour is also escaping the system along with the gases evolved and heavy water is 
continuously being added to the eiectrolyser to maintain constant level of electrolyte m the 
cell V I 

m p t 

For low tritium concentrations, all the tritium in liquid and gas phase will be in the 
form of DTO and DT respectively. Now the rate of increase in the number of DTO moles 
b the eiectrolyser per umt time, can be written as, 

I^Vrci “cto - ^ n tno^ V ^ ^ 

Where the term with’ (F) is the rate of addition of DTO in moles per second .being added 
with the feed, the term with n ]}T0 is rate of escape of DT with gas and the term with (V) » 

the rate of eschpe of DTO as vapour. 

The rate of escape of DT can be calculated from the knowledge of the electrolytic 
separation factor for tritium. Thus if r^ moles per sec. are electrolysed and is the 

electrolytic separation factor then for low tritium concentration a R can be defined as, 

(T/D) liq 

“e = TT7T3) s .. ® 

The number of DT moles escaping per second will be, 


where x is t 

lie fraction of DTO in the eiectrolyser at time %. 

The rate of escape of DTO as vapour can be calculated from the knowledge of the vapour 


A 10 

pressure, P v of heavy water at the temperature of electrolysis, pressure P in the electrolyser 
and flow rate, F of the gases evolving out of the electrolyser and the separation factor a 

for isotopic fractionation of tritium due to evaporation. Thus, 

Ht^TO^) = -p _ P V F 



where F g is the flow of gases from the electrolyser in moles per sec. 

The term due to the feed of fresh DjO to the electrolyser can be written as 

w JUf 

i n ™ (F) - h • * 


#• ‘ 

where F L is the feed rate of heavy water in moles per sec. and a is the mole fraction 
DTO in the feed. Equation can now be written as ■"* 

at nDT0 

n . 

x P v 


E ' 

% p 

. F . 

. 8 ' 

( 6 ) 


Also if the electrolyser contains N moles of heavy water at time t then, 

X 22? 




^ % 

dx 1 . d n„ 


DTO . d N, 




n DT0 

dx 1 

ar = ?r 

Substituting for d n DT0 from equation 6 in equation 9, 

d N 

at ‘ 







n , 

a . 

p x 

P Y g Of 

X. d N 

ar * 


( 8 ) 



This is the basic differential equation which can be solved for x knowing all the 
other parameters. Some typical cases for application to different operational conditions are 
considered in the following sections. 

Case 1: Diffusion Type Electrolyser with automatic Level maintainer: This case 
particularly refers to an electrolyser used by us (Milton Roy) in our experimental studies 
on cold fusion 4 . In this electrolyser deuterium diffuses through the palladium cathode and 
is delivered in absolutely dry condition. The level of electrolyte is automatically 
maintained. For such an electrolyser the total number of moles in the electrolyser, N t is 

constant and therefore N t = N and = 0. The term F g in equation 8 will be the rate of 


A 10 

evolution of oxygen and therefore, 

F =_ln (11) 

g 2 

Also the feed rate of heavy water will be equal to the total of the rate of electrolysis and 
loss of heavy water due to evaporation. Therefore, 

P , 

F = IL + llL . V fl2) 

L E 5 B P _ p , ^ ^ ' 

Equation 10 , now becomes, 



3t N 


P V 

1 + 2 ( P “ P v ) 

a • 


•w ' 


a E 2 °< 



i + 

2 ( p - p v y 



+ *«J P - Pv ) 





and rearranging equation 13, 

dx _ n E . dt 

— ~i — |TT- i i i i m »n , rn ’ 1/Mmt.ummmm 

A — B.x N 

jr „ 
<#' •#* 

Integrating for t=0 to t, we get 


a . A — B . xo 
a . A —B.x 

B . 

n . 

. t 


(17) - 

where xo is the mole fraction of DTO in the electrolyte initially. Equation 17 can be 
rewritten as, 

B . n. 

a . 



a — _x 0 

E . t 


( 18 ) 

This equation gives the mole fraction of DTO in the electrolyte at time t. As x, xq, and a 
have same units equation 18 can be applied for any common units of these parameters. 

It follows from eqn, 18 that x will increase or decrease with time depending upon 
the relative magnitudes of a and xp. Thus for xo > aA/B, the exponential term will become 
positive and x will increase with time; while for xq < aA/B, x will decrease with time. In 
both the cases the tritium concentration in the cell will try to attain a steady state value 
equal to the term aA/B as shown by the asymptotic behaviour of the tritium curves in the 
two cases (Fig. 2). 

If the tritium concentration of the starting electrolyte and that of feed water is same 
(a — xo) then eqn. 18 becomes, 


A 10 



B n 

1 > e- 

E . t' 



In this case also x will increase with time because it easily follows from equations 14 and 15 
that A/B > 1 rendering the exponential term negative. 

J ' 'S.'" 4& 1 ' 

Case 2: Batch Electroliers of non-diffusion type: In most of the elect^^ps experiments 
the cell employed is such that the mixed stream of deuterium and oxygen is coming out of 
the electrolyser and the cell is operated for a certain period without , adding fresh heavy 
water to it. In this case the first term in eqn. 10 will be zero ( as Fl = 0 ) ana also F g will 
now be equal to 3 n^ . Also will no more be constant and will beigiven as, 


N-(n I , + F 

* P -T & P 

m . t 


( 20 ) 



^*“^1 ife, P) 

N t = N - n E ( l + ) . t 

- n E ( 1 + l F^“P V ) 

Making these substitutions in eqn. 



P E 

simplifying, we get 

a E " 1 


n , 

% r '+ 


( 21 ) 

( 22 ) 

3 a v 

7 P - P v 
Integrating ai^4p^ting the condition x = xq at t = 0 



1 P v 

— ' p ~ 



o , 



L n - = 

a , 

. 3 __ 

+ 2 P 


1 + 

7 P 

-L n 

n . 



1 + 



Equation 24 can be employed to calculate the tritium enrichment during the electrolysis in 
terms of time of electrolysis. Alternatively, this equation can also be written in terms of Nt 




a , 





T P 



1 + 



7 F 

Using eqn. 25, tritium enrichment can be calculated knowing initial and final volume of the 
electrolyte. Equation 26 is m fact similar to a well known Rayleigh's equation given by 4 
which is as follows ^ 


( 1 + H 

r E . 1 


T _ 



— T 

1 - 

- x 0 ‘ 

i a E - l J 

Or —4. 


1 - 

- Xt 


va.<jP -si 

where H is the humidity of the gases evolving,. For low tritium concentrations xo and 
x t << 1 and therefore eqn. 26 becomes 


^ = ( 1 + H ) 


a , 



which is similar to eqn. 25 except* that it does not include the term due to isotopic 
fractionation of tritium due to evaporation. A plot of x/xg vs Ng/Nt as obtained from eqn. 
25 is shown in Fig. 3. / . • 

Let us now consid%p ; the different parameters appearing in equations 18, 24 or 25. 
are the electrolysis rate n , electr 

v, Sj 

factor a v and vapour "pressure of D 2 O, P v 

■fe*. . *• ^ * 

They axe the eiectrojylis rate n , electrolytic separation factor a jdisfciliatioE separation 

At E & 

The rate* of electrolysis n can be calculated from Faraday's laws of electrolysis 

knowing the current. The electrolytic separation factor Og depends upon the temperature 

of electrolysis and more so on the electrode materials. Not much data is available for 
electrolytic tritium separation factors. Generally a value of 2.0 is widely accepted 5 . 

The distillation separation factor er v can be calculated from the vapour pressure of 
D 2 0 and DTO.Thus, 


D 2 0 

D T 0 

The vapour pressures of D 2 O and DTG have been computed and are available in 
literature®. Table 1 gives the values of P (P v in eqns. 18, 25 and 26 ), P and o at 

i>2v I/TO W 

different temperatures 8 . 


A 10 








M Fleischmarm & S Pons, J Electroanal. Chem. Ml, 301 (1989). 

S B Jones et ai, Nature 338, 737 (1989). 

MS. Knshnan et a!, This Report, Paper Al (1989). , 

R, E. Treybal, Mate Transfer Operations p 308, (McGraw Hill, New Delhi, 1968). 

K M Kalyanam and S.K. Sood, Fusion Technology 14, 524—529 (1988), 

S. M.Dave, ’Tritium Separation Factors in Distillation and Chemical Exchange 
Processes', Report B.A.R.C. — 1168 (1982). 

... .la*, % 





P D 2 0 




0.02013 d 








0.06566 V 
















i .0 44210 






A 11 


K. Annaji Rao 

Chemistry Division ~ 


In nuclear fusion reactions involving D 2 helium may be one of the possible products. 
During electrolytic dissociation of D 2 O with Platinum/ Palladium electrodes if any fusion 
reaction 3 , 2 is taking place via the helium pathway with helium escaping to the gas phase it 
should be possible to detect and estimate the yield quantitative^- -Helium is reported to 
have been observed in excess of the background in the electrolysis experiments conducted 
at the University of Utah 1 . The number reported is of the order of 10 32 atoms per second 
(~ 0.013 /il per hour). In presence of large excess of D 2 and O 2 jfenerated during electrolysis, 
detection of He 3 / He 4 in the gas phase poses many problems. Among the methods available 
for the detection of low amounts of helium, gas chromatography (GC) arid mass 
spectrometry (MS) are normally preferred. Howeylh m this case mass spectrometry is 
complicated since helium is in trace concentration. and large excess of D 2 will interfere with 
the signal due to He 3 or He 4 ions unless fractional mass difference method is adopted. In 
order to concentrate the helium in the gas phase an experimental technique has been 
devised wherein D 2 and O 2 generated by electrolysis is catalytically recombined in situ 
facilitating gas collection over long periods of electrolysis. The results obtained by this 
technique followed by gas chromatographic analysis are given in this paper. Though GC/ 
MS analysis of the enriched sample may be more conclusive, a suitable GC/ MS gas inlet 
interface compatible with D«/ He is not commercially available 


The electrolysjg%eU v and the gas manipulation system is shown schematically in 
figure 1. This consists of a glass cell carrying aground glass joint with the cap connected to 
a mercury manometer with a 20 ml expansion bulb at the bottom, a silicone septum 
carrying vacuum tight electrode leads and a vacuum stopcock connecting the cell to a 
modified Toeppler pump with facilities to measure the gas pressure, volume and also to 
pressurize and transfer the collected gas to a syringe sampling manifold through a three 
way stop cock 3 . The third limb of this stop cock is connected to rotary vacuum pump. The 
catalyst for recombination of D 2 and O 2 at room temperature is a specially prepared 
platinum catalyst deposited on a thick synthetic fabric. It is freely suspended on the top 
inner side of the electrolysis cell. The catalyst has been independently assessed for H 2 / O 2 
recombination efficiency and found to have Tp 2 f° r H 2 reaction of less than 15 seconds in 
presence of sufficient (3 2 with gas volumes upto about 500 ml. The gaseous products are 
analysed by gas-chromatography using thermal conductivity detector either (1) on a 5 
meter x 3 mm i.d. molecular sieve 5A Column at 25°C with argon carrier gas which gave 
clear separation between helium and hydrogen for the analysis of trace amounts of helium 
in the sample or (2) on a 2 meter x 4 mm i.d. molecular sieve 5A column at 25°C with 
helium carrier gas for the analysis of H 2 (or D 2 ) O 2 and Ng in the sample. The lowest 
amount of helium detectable in the presence of large excess of hydrogen under the 
experimental conditions is about 0.01 pi (or about 1 ml sample with 10 ppm He). The 
second column provided a reasonably good analysis for the composition of the residual gas 
with respect to major constituents, Total volume of the electrolysis ceil and the gas 
manipulation system as well as the volume of each segment of the system has been 
determined to facilitate computation of the gas volumes and composition at any stage of 
the experiment. The electrolysis experiment has been carried out in three stages. In these 


A 11 

experiments 20 ml of 0.1 M LiOH or LiOD is used as the electrolyte. Experiments I and II 
are carried out with stainless steel Cathode and Anode mounted in concentric tubular 
configuration with 1 rnm electrode spacing. The electrodes are approximately 1.5 cm X 3 
cm a 0.04 cm in size. These experiments are used to test the system with respect to leak 
tightness., gas recombination efficiency, trace helium recovery and analysis. Experiments III 
to VI are earned out. with palladium/ platinum electrodes. In this case a palladium plate 
(I cm X 1.5 cm X 0.15 cm) is sandwiched between two platinum plates .0.%® X 1 5 cm X 
0.05 cm). 

Prior to starting electrolysis the entire system including^^lw cell with 21 
electrolyte is evacuated through the three way stop cock "tKoroughly degas cue 
electrolyte. After attaining vacuum the cel! is isolated firorrt the^ pump and rest of the 
system and the electrolysis started and continued for the required duration by connecting 
the electrodes to a D C,, power supply. At the end of the electrolysis the power is put off, 
system is allowed to stabilize till no further change in the mercury level of the manometer 
is noticeable. At this stage by suitable manipulatiorf*|>f the three stop cocks and mercury 
reservoir gas transferring, compression, measuremer^ ptipressure and volume and sampling 
for analysis are carried out. Whenever necessary, desired reactant gas is introduced info the 
electrolysis cell through the silicone septum .carrying the electrodes by means of a syringe. 
The electrolysis cell isolated under vacuuiu is^ found to retain vacuum for more than 48 
hours \arop in mercury level *0.5 cm),<Tt|,? results of the experiments with relevant details 
are given m Table .1. > 

Conclusions *Sk 

Pk # 

1) In contrast to the stainless steel electrodes Palladium cathode/ Platinum anode 
configuration resulted in pv deficient gas composition and Platinum cathode/ Palladium 
anode configuration showed Cp deficient gas In the initial stages of electrolysis much of the 
hydrogen (or D 2 ) may be absorbed by palladium reaching equilibrium in about 2.5 hours as 
indicated by the steady pressure build up. 

2) Though all the released gas is contained and concentrated to a residual volume of 1 to 3 
ml, no helium could be detected in all the experiments. 

3) Hydrogen (or Dj) absorbed on the electrode is slowly released under vacuum (note e). 

4) Platinum catalyst can be incorporated in the cel! design to recombine the evolved gas 
thus facilitating D 2 0 recovery and also possible recovery of any tritium in the gas phase 
back into the aqueous phase. 

5} In case an independent proof, for the cold fusion via the helium pathway is available, 
this method can be adopted suitably to detect helium and substantiate the finding with the 
possibility of unambiguous evidence obtainable from GC / MS analysis of the concentrated 
residual gas. 

Explanatory note to the Table 

(a) In case of I and JI electrolysis is carried out for varying duration, to ascertain, the 
viability of the method and recombination efficiency of the catalyst. 

iu ^ ^ these cases 150 /d of 1 % He in argon is added to the ceil prior to electrolysis and 
the residual gas at the end analysed for He. Recovery is found to be better than 90%. 

(b) (i) In case of Til to VI initial evacuation of the system is carried out with the D.C. 


A 11 

potential for about 5 minutes, then evacuation continued with the potential off for about 5 
minute, system isolated and the regular electrolysis carried out {system purgi ng) . 

(a) In these cases mercury level in the manometer showed a steady drop for the first 2.5 
hrs and then slowed down. 

(b* + ) In case III the residual gaa is used up in analysing for trace helium. Hence D2:02 
composition is not determined. 

(c) At the end of experiment IV the electrolysis cell is maintained in the isolated mode 
overnight. The mercury level dropped by about 2 cm overnight and the residual gas on 
analysis showed 90% Dj indicating release of D? under vacuum from the electrode. 

Acknowledgement / V 

The author is obliged to Dr. N.M. Gupta and A„I|| Belapurkar of Chemistry 
Division for providing the platinum catalyst. The author, is alsb thankful to Dr, H.M. Iyer, 
Director, Chemical Group and Dr, J.P. Mittal Head # 0$tlnistry Division for their keen 
interest and helpful suggestions during the course of thWAsfork. 


Martin Fleischmann and Stanley Pons, J. Blectroanal. Ohem. 261, 301-308 (1989). 

C.K. Mathews, G. Periaswami, K.C, Srimvas, T. Gnanasekaran, S. Rajan Babu, 
C. Ramesh and B. Thiyagarajan. Indian J. of Technology 27, 229—231 (1989). 

K.A. Rao, R M. iyer, Indian J. pf Technology 16, 44—45 (1978) 


\ c 





D, G A S' «jLO A DING 





P. Raj*, P. Suryanarayana* A, Sathyamoorthy*, and T. Datt&t 

•Chemistry Division 
t Radiochemistry Division. 


The possibility of D— D nuclear fusion in some deutenum-snetal systems, uuuct 
ambient conditions, has aroused feverish world wide interest. Most of the work reported, so 
far, concerns deuterium charging of Pd metal through electrolysis oft D 2 Q 

In Chemistry Division, we have carried out some experiments on the deuteriding 
behaviour of Ti metal, through gaseous route, in theffibgorption as well as desorption 
modes, with the view to look for the fusion product^; .niutrons m the present case. This 
kind of experiments have been reported by FrascaDi Group in Italy 1 . These authors 
detected neutron emission lasting over a period oftsfev& hours. 


r -% 

Experimental arrangement, for distending Ti metal is shown in Fig. 1, which is 
self-explanatory. This set up has beep«rputinely used for high pressure hydriding studies on 
several systems, reported by us 2 " 4 some of the experiments reported here, deuterium 
pressure was cycled between high^dTbw values by simply changing the temperature of 
the cell housing the sample. Most of the experiments were done in the desorption mode. Ti 
metal pieces (cut from a jheej/'ypsre surface cleaned and subjected to activation treatment 
before Da loading and subse^uftit desorption treatment etc. 

Neutron set up consists of an array of 24 He 3 counters arranged in a well 
like geometry, These counters (each 50 cm in. length 2.5 cm in diameter and filled with He 3 
at 4 atm.) housed lift paraffin moderators, are all connected in parallel to a single 
pre— amplifier . The counting efficiency of this system was found to be «10%. The counts are 
recorded in 8|02 channel multi— scalers. In the experiments reported here dwell time of 
40 sec. was fixed, so that each point in Fig. 2(a) to (d) represents the number of counts per 
40 sec. The background counts collected for about 10 days, before the start of these 
deuteriding experiments, was found to be quite steady 60 counts/ 40 secs. This background 
count rate continues to be the same well after our experiments. 


In the first set of experiments, starting from 3rd June 1989, after activating Ti 
metal pieces, D 2 gas was contacted with the sample at a pressure -10 atm. while keeping 
the sample at low temperature (-77K). After a soaking time of -20 min., sample 
temperature was raised gradually, while simultaneous evacuation was started. Within 
about 15 min. the neutron counter registered an increase in count rate reaching a max. 
3900 (as compared to back-ground counts of -60), see Fig. 2(a) On withdrawing the 
reactor from the counting well, a considerable reduction in the counts was observed. On 
re-introducing the reactor after background counts are restored, an additional peak like 
structure was observed. Although the evacuation was continued, no further increase in 
count rate over the background could be observed over the next twenty hours. 


B 1 

Next experiment on the same charge was carried out by repeating the conditions of 
first experiment. The results of this experiment, dated 4th June 1989, are shown in 
Fig.2(b), Again two peak like structures, each lasting for about 30 min. and separated by 
50 min, were seen. However, the intensities of both these structures are greatly reduced, as 
compared to the first experiment (3rd June 1989), the max. counts being ~ 700. 

In the third experiment, with the same charge of Ti pieces, D 2 ga^tessure was 
made to cycle between -50 atm. to -13 atm. by changing the reactor temperature from 
room temperature to 77 K. In this case large changes in counts, as a function of time were 
noticed. An increasing trend of counts initiated at -2330 hrs on 4th June 1989 lasted for 
almost 7 hrs. with an estimated integral counts -6.5x10 s . Even atfier this long bursts like 
structures, some additional peaJks were observed on 5th June 1989. With no further 
structures observed over the next few hours, desorption was earned out after loading the 
sample with D 2 gas with the sample temperature at -77K. By raising sample temperature 
gradually, while simultaneously evacuating, a much bigger’ structure lasting for -2 hrs. 
(from -1830 to 2030 hrs. on 7th June 1989) was seen Fig 2(c). An approximate estimate of 
integrated count over this period is 7x10 s . Further experiments with this charge, involving 
~2 >°dowed by prolonged periods of evacuation at temperature upto a max. of 

~200 !J C, did not show further structures. 


Second series of similar experiments on a fresh charge of Ti from the same source 
did not exhibit exactly similar behaviour, astound for the first charge. However, one set of 
experiments on lith June 1989 see Fig. 2(d) involving pressure cycling, followed by 
evacuation, exhibited increase in count rate lasting over a period of -100 min. In this case, 
the scatter in the counts was found to be rather large and maximum counts upto 10 s / 40 s 
were observed, as compared to background count of -60/ 40 sec. This charge showed no 
further increase in count rate even after various treatments. 


further experiments are planned — (i) to study all possible parameters relating to 
the observed increase in the count rates, (ii) to identify the source of these extra counts, 
and (iii) to investigate the energy and time structure of the radiation responsible for the 
observed peak like structures. 


Authors are indebted to Dr. R.M. Iyer, Director, Chemical group and Dr. J.P. 
Mittal, Head, Chemistry Division for helpful discussions and encouragement. Sincere 
thanks are due to Dr. P.R. Natarajan, Head, Radiochemistry Division; Dr. H.K. 
Sadhukhan, Head, Heavy Water Division; Dr. C.K. Gupta, Head, Metallurgy Division and 
their colleagues for providing neutron counting facility, Deuterium gas and Titanium 
metal, respectively. 


1 A. De Ninno et al, Europhysics Letters 9, 221 (1989). 

2 P. Raj, A, Sathyamoorthy, P. Suryanarayana and R.M. Iyer, J. Less-Common Metals 
123, 145 (1986). 

3 P. Raj, A. Sathyamoorthy, P. Suryanarayana, A.J. Singh and R.M. Iyer. 
Less— Common Metals 130, 139 (1987). 

4 P. Raj, P. Suryanarayana, A. Sathyamoorthy and R.M. Iyer, Mat. Res. Bull. 
24(6), 717-724 (1989). 







V. K. Shrikhande* and K.C. Mittal t 

B 2 

♦Technical Physics and Prototype Engineering Division 

$ Plasma Physics Division ' 


Cold fusion experiments were initiated with solid targets made from titanium loaded 
deuterium gas on receipt of reports of the successful Frascati experiments 1 . The 
absorption of deuterium by Ti is a reversible process and titanium is heated in a 

deuterium atmosphere, the reaction will continue until the concentration of deuterium in 
the metal attains an equilibrium value 2 . This equilibrium value depends on the specimen 
temperature and the pressure of the surrounding deuterium atmosphere. Any imposed 
temperature or pressure change causes rejection or ajlsorption of deuterium until a new 
equilibrium state is achieved. If the surface of ‘4jtatiium is clean, the rate of absorption 
increases rapidly with temperature, At temperatures above 500 °C, the equilibrium is 
achieved in a matter of a few seconds. However deuterium absorption is considerably 
reduced if the surface of Ti is contaminated 'with oxygen. Keeping in view these facts, a 
procedure was evolved for titanium target preparation and subsequent deuteration. The 
following sections describe the detaijs qf preparation of the targets, their chemical cleaning 
and degassing followed by deuteration process. 

Preparation of the Targets 

Titanium targets of different, sizes and shapes (planer, conical etc) were prepared. 
1 argets were typically a fraction of a gram in mass and were machined out of a Ti rod 
using tungsten carbider tools with continuous cooling arrangement. Care was taken to avoid 
overrating during .machining because any overheating could harden titanium and thereby 
inhibit its capacity -to absorb H 2 /D 2 . 

The machined targets were first degreased ultrasonically in trichloroethylene. Then 
the oxide layer if any was removed by immersing the targets in a 1:1:1 mixture of water, 
nitric acid and sulphuric acid. They were then rinsed in water and dried in acetone. This 
was followed by HOI treatment to form an adherent hydride layer on the surface. Targets 
thus prepared were preserved in a moisture free environment prior to deuterium 

Degassing and Deuteration of Targets 

The chemically cleaned targets were first degassed by heating to ~90G °C in a glass 
vacuum chamber using a 3 Kw, 2 MHz induction heater. Degassing was continued till a 
vacuum of less than 10" s Torr was achieved. Targets were then heated to ^600°C in H 2 
atmosphere at a few Torr pressure and allowed to cool. H 2 was absorbed in the targets 
while cooling. Absorbed ll 2 was released again by heating to 900°C. At least three cycles of 
Il 2 absorption/ desorption were given to create active sites for D 2 absorption. 

After release of all H 2 , the targets were heated to ^600 °C in D 2 atmosphere at few 
torr pressure and allowed to cool by switching off the induction heater. D 2 gas was 
absorbed while cooling. At least three cycles of D 2 absorption/ desorption were given, 
similar to H 2 absorption/ desorption. Fall in pressure recorded by an oil manometer is a 


B 2 

measure of the quantity of D 2 absorbed It was found that the quantity of gas absorbed 
increased m each new cycle and tended to saturate in the 3rd or 4th cycle. Table I 
illustrates the maximum absorption of hydrogen and deuterium in different Ti targets. 

^ It was noticed that targets could typically absorb sdO 1 ® molecules of D 2 , 
Considering that mass of Ti is a few hundred milligrams, this correspond|^%’ an. overall 
D/Ti ratio of slO -3 only. However, if most of the absorption is restricted surface, as 

we suspect, it is likely that the D/Ti ratio is higher than 0.001 in the neeg-^Jface region. 

While preparing the targets, we found that successful deuter ati on depends on 
various experimental factors as listed below: r 

0) biitial sandblasting of the targets for cleaning and rqgghl:t|iS| of the surface leads to 
better absorption of D 2 . (it) Impurity content (such as 0 2 , N 2 etc) in D 2 should be <0.1%., 
(hi) Since the glass vacuum chamber is isolated fro, pa tip pumping system during D 2 
absorption, it is important that the vacuum chambcfilbp leak tight. Small air leaks may 

contaminate the D?. 

* t# I 

The deuterated targets were sent t % the iseutron Physics Division for analysis in 
quest of evidence for cold fusion Ref. 3 athAAdescribe the autoradiography and neutron 
counting results. ^ 


As mentioned earlier the 4?' (iteration of the titanium targets was carried out using a 
3 Kw induction heater operating at 2 MHz frequency. The power supply of this heater 
became defective in July 8$ following failure of the main driver tube. Since then gas 
loading of targets coijld J$it be carried out in this division. Similar experiments were 
thereafter commenced at the Heavy Water Division using a resistance furnace as described 
in Ref. 5, However although the loading procedure adopted there was such that very large 
quantities of D 2 gas/to:6 litres at 1 Kg/cm 2 ) could be successfully absorbed in titanium 
pieces (mass rd>. grams), none of the Ti samples have shown any evidence of tritium so far. 
It is possiblffl|haf. pse of high frequency (2 MHz) induction heating may have had some role 
in causing .the detect able levels of cold fusion. 

When" a metallic object is heated by induction heating, the current distribution 
within the object is non-uniform with the current density decreasing exponentially from 
the surface to the centre of the metallic work load 6 . The characteristic penetration or skin 
depth 6 is defined as that distance over which the current density is reduced to 1/e times 
the surface value and is given by 

5 = (/?/firp)h 2 

where p is the resistivity, /i, the permeability of the workload and f is the frequency of the 
applied alternating magnetic field. For a 2 MHz induction heater the skin depth in 
titanium works out to be ~0 1 mm. It is believed that most of the absorbed D 3 gas is 
accumulated in the near surface region even though the entire sample would have reached 
high temperatures due to conduction. Hence it is likely that D 2 density is very much 
higher in the near surface region though the gross .D/Ti ratio is hardly 0.001. 

Further investigations to confirm these conjectures axe underway. 


B 2 


The authors wish to acknowledge the significant spade work done by Dr. S.K.H, 

Auluck, in arriving at the optimum conditions for loading of H 2 /D 2 in machined Ti targets. 
We are also grateful to V ,G. Date of the Atomic Fuels Division for fabrication of machined 

•&T oKSSfA, 


1 . 

2 . 




6 . 

A De Ninno et al, Europhysics Lett. 9, 221 (1989). 

Preparation of Hydride Configurations and Reactive Metal Sux 
Patent Report U.S A. 6-611, 773 (1985). 

R K Rout et al, This report. Paper B3 (1989). 

A Shy&m et al, ibid, Paper A3 (1989). 

M S Krishnan et al, ibid, Paper B4 (1989). 

E.J. Davies and P.G. Simpson "Induction Heating Handbook" 
Me. Graw— Hill, (1979). 


■ ^ 

Maximum Absorption of H^/D^ in Different Tit anium Targets 

Silver G.L. U.S. 

New York, 

Sr. No 


Mass 4 . ~ 

"■— * — % — 1 

II 2 Absorption 


(g),T~ * 

mm of 

mm of 

mm of 

ram of 








































































80. 0 





























2 JO 




































R K Rout, M Srinivasan and A Shyam 


Neutron Physics Division 

For the last few months, hectic activity is underway in various' laboratories to study 
the Cold Fusion phenomenon. De Ninno et a! 1 reported emission of neutrons from titanium 
metal loaded with deuterium gas under pressure. Similar experiments have been conducted at 
Trombay. We report here evidence of cold fusion in D 2 gas loaded Ti and Pd targets through 
the use of autoradiography for spatially resolved detection of tritium. Our study employed 
three different techniques to observe tritium: 
fi) Autoradiography using X— ray films. 

(ii) Characteristic X— ray measurement of titanium, excited by the tritium 0. 

(iii) Liquid scintillation method for tritium 0 counting.! 

Loading of Deuterium * ,"Cv X 

■sir -w 

Titanium and palladium metal samples of various shapes and sizes were loaded with 
deuterium by two different ways. In the first method 2 an individual titanium target was 
heated by R F heating up to a maximum temperature of £U 900 °C in vacuum and then in 
deuterium gas atmosphere to absprb deuterium. In the second method 3 the foils of palladium 
(Pd-Ag alloy) were heated (by ohmic' heating) up to a temperature of 600 °C in vacuum 
(1 Q~ 5 ibixi of Hg) and then, in Dg gas. The deuterium gas used for loading jiad a tritium content 
of <5,5X10' 4 Bq/ml of gas. corresponding to a T/D ratio of ~4X10‘ H . 

Jr £» 


Autoradiography is a simple and elegant technique of detecting the presence of 
radiation eputtm'g ' zones. This technique has the advantage of being free from any 
electromagnetic-interference (pick ups, discharge pulses etc), has relatively high sensitivity as 
it can integrate over long exposure tunes and can give very useful information in the form of 
space resolved images. In order to achieve good resolution of the image, the sample was kept 
very dose to the X-ray film. Standard medical X-ray film of medium grain size (10 to 15 fi m 
in diameter) on cellulose triacetate base was used for this purpose. The exposure time used for 
the deuterated samples varied from 18 hours to a few days. At times a stack of several films 
was used. In some cases films were placed on both sides of the sample. For latent image 
formation we used IPC (India Photographic Company Ltd.) made 19B developer and IPC 
made fixer. The developing time was typically 4 to 5 minutes. Out of many samples which 
had absorbed D 2 gas, only a few showed a latent image. The results axe tabulated in Table I. 

The radiograph (Fig.l) of some of the deuterated titanium disc targets showed several 
spots randomly distributed within the sample boundary. The occurrence of spots all along the 
rim of the machined target is very intriguing. Repeated measurements over a period of one 
month, with the same sample with varying exposure times gave almost Identical pattern and 

.*.•.**. _ ,i! it. .a i.i *?• . . * *» . * 

well entrenched in the face of the titanium lattice. The fact that the second film of a stack of 
films exposed to the target also indicates similar though less intense spots, rules out the 
possibility of any kind of chemical reduction reaction caused by the deuterium or hydrogen in 
the target being responsible for causing the spots. The X— ray image (Fig. 2) of a conical 
target showed a diffused projection of the cone. 



The .image of Pd—Ag foils (Fig, 3) however exhibited a more uniform image. The 
images however indicated variation in intensity and some spots but on the whole the fogging 
was more or less uniform. Unlike deuterated titanium targets, the intensity of fogging of 
deuterated Pd foils reduced very rapidly i,e. within a couple of days the activity reduced 
below measurement level. 

Measurements of X-ray Emission 

The characteristic X-rays emitted from the deuterated metals (TP and Pd) were 
studied with the help of a Si (—Li) (Silicon — Lithium drifted) detector JyTfS Nuclear Physics 
Division. The detector had a beryllium window of 75 pm thickness. The X—rays of Ti (Ka = 
4.5 Kev, K0= 4.9 Key) were observed in case of conical (Fig.4) an<fli£c (Fig.5) samples. The 
count rate of the conical sample was much more than that of the disc sample. Some of the 
deuterated Pd—Ag foils indicated the X-ray peaks (Pig. 6) corresponding to titanium 
presumably because of a small amount of titanium impurity picked up by the foils from the 
D 2 loading chamber winch had earlier been used for loading of Ti samples. We did not observe 
the t X—rays of palladium or silver. 

fTj F || 

Liquid Scintillation Counting 


This was carried out at the Health Physics Division using the facilities described in 
Ref. 4, The sample was simply dropped into a vial containing liquid scintillator cocktail and 
the tritium activity was counted by two photomultiplier tubes in coincidence. The typical 
activities were 50 to 1000 Bq as compared to a background of less than 0.2 Bq. No correction 
was applied for possible quenching/ shadowing effects. 

Results and Discussion 


, . The fogging observe^ m autoradiographs (Figs.l, 2 & 3) is the combined effect of 
tritium betas and characteristic X—rays of the host material. The radiograph of the disc 
sample (Fig. 1) indicates, .evidence of tritium localized in the form of microstructures. These 
spots are unevenly distriSuted on the face of the titanium, there being about 60 to 70 spots in 
all. On correlation with the X— ray counts under the peak (K X— ray peak) and liquid 
scintillation counting results it was found that the each emitting spot corresponds roughly to 
10® to 10 10 atoms of tritium, In comparison the total number of deuterium atoms loaded in 
the disc sample was 10 19 to 10 2Q . The X— ray images (Fig. 3) in case of Pd—Ag foils were 
uniformly fogged and intensity of fogging reduced very rapidly with time unlike with 
titanium. This type of loss in image may be attributed to the high mobility of tritium in 
palladium as compared to that in titanium. Observation of K X-ray peaks of titanium 
(Figs. 4, 5 k 6) by Si (—Li) detector was the result of excitation of K -shell by tritium 0, L 
X-ray of palladium 3.6 Kev) or silver (2 3.8 Kev) was not observed because of low 
Increscent yield for t X-ray and the detector window being too thick (75 am) to allow 


on the surface of the samples exceeded the total quantity of tritium initially contained in the 
deuterium, gas used to load the samples and hence the gaseous tritium, even if preferentially 
absorbed by the samples cannot explain this phenomena, Undeufcexated metallic targets 
machined out of the same titanium rods did not indicate any detectable tritium, ruling out 
any contamination pick up during target fabrication/ handling. 

Summary and Conclusions 

The evidence presented in the paper seems to be indicative of cold fusion reactions 
occurring in some of the deuterium loaded titanium and palladium targets. It has not been 


possible to conclusively establish whether the fusion reactions occur during the deuteration 
process or subsequently. Also it is not clear whether the reactions occur in sporadic bursts or 
continuously. However one of the disc targets, which gave impressive spotty radiograph did 
give rise to a significant neutron burst which produced 1Q« neutrons 5 over a period' of .85 
minutes. r 

Acknowledgements ^€#*| 

RADnI he i aUt i h ° rS 8incerely wlsh to ex P ress the » r gratefulness to Dr-. P K Iyengar Director 
fuc V-i keen interest and constant guidance in the presentwoxk; We are also thankful 
to M S Knahnan, S K Malhotra, S Shnkhande and K C Mittal for supplying the deuterated 
targets. W e also gratefid to Drs^ \ S Ramamurthy and Madan Lai for the X-ray spectral 
measurements. The authors would like to express their thanks to 'Dr. T S Iyengar for carrying 
out the liquid scintillation counting of the targets. 




.. ' ' . 

f 4 

A De Ninno et al, Europhysics Lett. 9, 221 (3MSL/ 

E 9 ^ Srikhande, This Reporf^iper B2 (1989). 

M S Knahn&n et al, ibid, Paper B4 (198®/C\ ' 

T S Murthy et al, ibid, Paper A9 (1989):. > 

A Shyam et al, ibid, Paper A4 (1989). 


Sample No, 

_ TJDS 001 


PS001 , 

Material A ... 

* ' Ti 


Pd— Ag 

Shape of Sample 




Sample Mass (mg) 




D 2 Absorbed (mg) 




D 2 Loading process 

Ref .2 

Ref .2 

Ref. 3 

Date of D‘j Loading 

14 — 6 — 89 



Date of Exposure 

23 — 6—89 



Ejtposure Duration 

Figure No. 

66 Hr 


24 Hr 


88 Hr 


No. of Times Repeated 




Si(— Li) Result (Bq) 




Date of Measurement 




Total Tritium Atoms 

—1.5 10« 

266 1 G» 

2 l-§ 10 52 

T/D Ratio 

sd .2 10 "$ 

23.2 10-8 

2 ? 1Q-S 

Liq— Scm. Result (Bq) 




Date of Measurement. 




‘Estimated from titanium (present as impurity) X— rays; can be inaccurate. 


B 4 


M.S.Krishnan, S.K.Malhotra, D.G.Gaonkar, V.B.Nagvenkar and 

H . K . Sadhukhan 


Heavy Water Division 

After the first announcement reporting the observation v # r cdld Fusion 1 ' 2 further 
evidence supporting the same has appeared in scientific literature although many other 
groups have failed to obtain positive results. Palladium?' and titanium loaded 
electrolytically 1 " 2 and titanium loaded directly with deuterium gas^ have been reported to 
emit neutrons. Interestingly gas loading experiments involving Pd— D have not been 
reported so far. Such experiments were therefore conducted recently in our group. Tritium 
measurements in gas loaded Pd-D targets have been - carried out. The present paper 
summarises the results obtained so far to ascertain whether cold fusion reactions occur in 
gas loaded Pd targets also, 4“^* 

Preparation of Pd Samples 

For loading deuterium gas in Pd, two types of samples were used. One was Pd-Ag 
alloy supplied by M/s. Johnson— Mathey and was directly used without further surface 
treatment. The other type of sample used was Pd— black powder which was prepared from 
PdClj. Absorption of deuterium dbjr Pd— black was very fast and readily gave a 
stoichiometry of D:Pd = 0.6 as deduced from the drop in gas pressure during loading. But 
in case of Pd— Ag ailoy*th| "absorption was rather slow and also the amount of deuterium 
absorbed was much less than that corresponding to a D:Pd ratio of 0.6. 

D 2 gas Loading Procedure 

The Dj gas used for gas loading was prepared from D 2 0 procured from Heavy Water 
Plant at Barpda fGujrat state). This D 2 0 had a tritium activity of 0.075 nCi/ral, The gas 
generated from it by reducing with Na in a vacuum system under stringent conditions, was 
stored in a s.s. cylinder under pressure and liquid N 2 cooling in the presence of activated 
charcoal. The D 2 gas thus produced was not further analysed for tritium as it we® 
expected to contain not more than 0-038 nCi/1 activity. This corresponds to a (T/D) 
isotopic ratio of 3X10" 14 . 

The schematic drawing of the experimental set up used for gas loading is given in 
Fig, 1. It essentially consists of a vacuum system equipped with a rotary pump and an oil 
diffusion pump giving a vacuum of 10' 6 torr. The s.s. reaction vessel C containing the Pd 
sample is connected to the vacuum system through an s.s. buffer tank B. Deuterium 
cylinder D is connected to the vacuum system through needle valve VI. The system is also 
equipped with a pressure gauge G and a manometer/ pressure gauge VG. The entire 
system was tested for a vacuum of 10" 6 torr and pressure of 100 Kg/cm 2 . A weighed 
amount of Pd— black or Pd— Ag alloy was taken m the vessel and heated to 600 °G for 2 
hours under a vacuum of better than 10~ s mm. After cooling to room temperature, 
deuterium gas was filled at 1 atm. pressure and the system sealed off to attain equilibrium. 

After completion of gas loading the vessel containing the Pd sample was isolated 
from the filling system and transferred into a closed glass container in a dry enclosure, free 
of moisture or oxygen and kept for equilibration for several hours. Adequate precautions 


were taken to avoid inadvertent exposure to moisture since this would, lead to catalytic 
recombination of absorbed D| gas with 0 2 , accompanied by considerable increase in 
temperature. The entire deuterium absorbed in Pd would be lost if dire precaution is not 
taken. This was confirmed in one set of experiments where the loaded sample was 
accidentally exposed to ait and. fcfao resultant water sample aquilibrated did not show any 

Tritium Analysis . ' J f 

4 w* M 

In oid, er to measure the tritium if any produced due to cold fusion reactions the 
deuterated P d samples were kept in contact with distilled water for a few hours to extract 
the tritium by isotopic exchange into the water. The water samples containing tritium 
were later sent to the Tritium group of the Health Physics Division for liquid scintillation 
counting using procedures described in this report elsewhere. 

In converting the measured tritium activity in the distilled water to a calculated 
tritium activity originally present in the Pd samples, fbereiis a conversion factor to be used 
which is computed as follows: 

Taking into consideration the exchange reaction 

T & b S + H2O = H^bs + HTO 

( 1 ) 

and applying laws of Chemical Equilibrium, one obtains the following relation for low 
tntium concentration : 

nt[x # (i/K -p i) - Xjj 

% k, \ refer t,o tritium atom fraction in the absorbed and the liquid phases 
respectively and subscripts T-and e refer to initial and equilibration conditions respectively. 
% n| are § m of gas absorbed in metal and gm moles of water taken for 

equilibration, h is. the equilibrium constant for reaction (1) and was taken to be same as 
for exchange ofjtptium between hydrogen and water, since at equilibrium the system 
consists mainly of H both in the absorbed and the liquid phases. The value of K is 4 6.128 at 
30 0. Yi calculated from equation (2) gives the number of tritium atoms produced for 
every deuterium atom absorbed in the metal. The tritium atom fraction X can be 
calculated from the tritium activity A applying equation 3. 


d x 3200 

where A is activity in Ci/ml and d is density of water in gm/ml (d = 0-996542 at 2T°C). 
Results and Discussion 

Table I summarises the results of the tritium measurements on the four samples 
studied by us. The first three rows give the experimental parameters viz. mass of sample, 
volume (oTP) of deuterium absorbed and the volume of water employed for isotopic 
exchange. The fourth row gives the D/Pd ratio. The fifth row shows the time for which 
the gas loaded Pd. samples were kept undisturbed in the same pressure at which Dj gas has 
been loaded. The seventh row .gives the tritium activity of the water after equilibration. 
JlllS list tw O' TOW'S tflY* •*#!»«* vv£ & * ' » ■* * * 

re the idea of absolute amounts of tritium present in the metal before 


B 4 

equilibration. The T/D ratio is about two-three times more in the case of Pd— Ag foils as 
compared to Pd black (6th row). It should be noted that the T/D ratio in the targets is in 
the range of 10" !2 to 10“ u which is more than two orders of magnitude higher than that in 
the initial deuterium gas used for loading. 

A fresh lot of Pd black (10 gms) and Pd— Ag foils (0.43 gms) were equilibrated- with 
gas and after this accidentally they were exposed to air . These became intensely hot 
and on equilibration with distilled water and analysis of T in the equilibrated water, the T 
activity observed in both was below detectable limit. The same ;i sartflfc4)f Pd— Ag foil 
(0.43) was reactivated (660 °C, 10~ 5 mm) and again loaded with D 2 gas which shows the 
results as in the 3rd vertical column, proving that the residual in Pd is nil. It is 

also seen that when the equilibration time is more, more tritium production is observed. 

The method of isotopic exchange for extraction of from metal to water phase 

is less cumbersome and the T/D ratio obtained from .equation (1) gives at least a 
conservative estimate of tritium produced. 

Autoradiography of Samples 

The Pd— Ag foils were also s ub jec t ec^fP autor adi ogr aph y to obtain images of tritium 
distribution as described in detail elsewhere in this report 5 It essentially consists of 
keeping the gas loaded sample on x— ray film and developing it after allowing adequate 
exposure time. The Pd— black samples co-uld not be subjected to Autoradiography as the 
powder got stuck to the film. But in case of Pd— Ag foils it gave unmistakable fogging of 
x— ray film corresponding to the geometrical shape of the foils, thus indicating the emission 
of some radiation from these foils and the only radiation being emitted from these foils 
which can be thought of is the $ -rays from tritium and the characteristics X-rays of the 
metal excited be tritium, 

Summary and Conclusions# 

A. '^W 

Gas loade„d y P<t' samples have provided evidence for the first time of the presence of 
tritium, strong|;^|suggesting the occurrence of cold fusion reactions. The Pd— D system does 
not require Tligh, pressure of D 2 gas and also no external perturbation is required to create 
non equilibrium conditions as suggested by De Ninno et al 3 . Although no quantitative 
comparisohllcan be made between electroiytically loaded and gas loaded Pd experiments. 
The present results confirm that electrolysis is not the only approach to inducing cold 
fusion in Pd lattice. Unfortunately no neutron measurements were carried out in the 
present work. A correlation between neutron production if any and tritium production 
would contribute significantly towards understanding the mechanism of cold fusion. 


Authors sincerely wish to express their gratefulness to Dr. P.K. Iyengar , Director, 
B ARC for his keen interest and constant guidance in the present work. We are thankful to 
Dr. M Srinivasan for critically reading the manuscript and offering ,many suggestions. We 
also thank Dr. A Shyam and Mr. R.K. Rout for carrying out the autoradiography of the 
samples. This work would not have been possible but for the unstinted cooperation of Dr. 
T.S. Iyengar, of the Health Physics Division, who carried out the Tritium counting of all 
the water samples, The authors also wish to thank Shri. V.H. Patil, Heavy water Division 
who has been of great assistance in the experimental work. 


dN- 03 

B 4 


M Fleischmaan And S. Pons, J. Electroanal. Chem. 261(2 A), 801—308 (1989). 

S.E. Jones et ad, Nature (London) 338, 737 (1989). 

A De Nirmo et al, Europhysics Letters 9, 221(1989). 

SM Dave, S.K Ghosh and H.K. Sadhukhan, 'Tritium Separation factors m 
Distillation and Chemical Exchange Processes', BARC Report BARG— 1168.(1982). 

5 R K Rout et al, This Report, Paper B3 (1989). 


Summary of Tritium Measurements 


Nature of Sample 

Pd black powder 

Pd%V^ foil"' 

.♦ % % 

Pd Ag foil* 

Pd Ag foil? 
(single foil) 


Mass (gn») 

20.0 / 

t >*0.96 




Volume of Dj gas 






D/Pd ratio 






Time of equili- 
brations (hrs) 

i «£#'*# 
lw M 
(T % 

Sfc % 





Vol , of water used 
for isotopic 
exchange (ml) 

«/ .50 






Tritium activity 0.22 

of water after 
equilibration (nOi/mi) 





(T/D) rati^C'-- 
in Pd 

3.24xl0' 12 

1.08xl0" u 

8, 3 lx IQ* 1 5 

8.67xl0" 12 


Absolute tritium 
activity (nCi) 



4 32 



Number of tritium 
atoms in Pd 

2.3lxlO n 


8.96xlG ll > 

2 4xlO S! 

•These were the triangular foils also studied by Autoradiography 

♦This was a single 11.5 cm dia foil; same as cathode used in cell of paper A2 


sr ! 


T o Vacuum 





R. Chidambaram and V. C. Sahni 
Physics Group 

In a recent electrochemical experiment with palladium cathode and plat mum anode, 
immersed in 99.5% DjO + 0.5% H 2 0 plus 0.1 M LiOD, Fleischmann and Pons 1 claimed to 
have seen an ’excess heat', which they ascribed to 'cold fusion 1 of deuterium nuclei 
electrochemical! y infused mto the palladium lattice. These authors arid lopes et al 2 have 
attempted detection of neutrons/ 3 H as signatures of fusion based on -the well known 

i ' 

D + D - 3 He (0.82 MeV) + n (2.45 MeV) t . 

D + D -4 3H (1.01 MeV) + p (3.02 MeV); 

and have found some favourable evidence. So far efforts" to Confirm their findings made in 
various laboratories, including BARC, have been somewhat inconclusive, although some 
neutrons appear to have been seen occasionally. The reported neutron production 1 is about 
six to nine orders of magnitude less than what the ‘excess heat 1 would imply. These 
observations have generated a great amount of debate 3 concerning the nuclear physics of 
D-D reaction; novel ideas have been proposed including the reaction D + D -* 4 He, with 
the added requirement that the energy (.23.8 MeV) be delivered directly to the lattice. 
Even when they have been observed, ofttndt has been claimed that neutrons appeared only 
intermittently. In view of these feat ures^we’ deem it prudent to draw attention to some of 
the relevant solid state aspects of P4— D;||stem which might be involved in some way. 

Absorption of hydrog^tf/bdeuterium by Pd has been studied for a long time and is 
known 4 to be strongly exhtfietpiic. Neutron diffraction 5 experiment shows that H/D goes 
mto the octahedral sites in. the Pd lattice. Accompanying this uptake is a structural 
change, with the solid ; changing from the a phase at low H/D concentrations to a + (3 
phase at higher va|ue&; to an eventual (3 phase. Both the phases are f.c.c., with cell 
constants 3 89 A 0 and 4.03 A 0 respectively. Calorimetric studies 4 , using activated Pd and 
molecular D 2 gas, sh.dw that, at 30° C, the heat released during the formation of PdD* rises 
from ~ 7.50 kcals per mole of D 2 for x « 0.023 to ~ 8.43 kcal per mole of D 2 for x « 0.422, 
with the authors asserting that ‘these heats show a definite increase for each increment of 
gas added throughout the mixed phase region 1 . Thus as more D atoms are loaded into Pd, 
one may expect that the (3 phase regions (with a higher lattice constant and presumably 
better cohesion) would grow in the matrix of a phase, leading to regions of localized 
strains. We may conjecture that eventually abrupt atomic readjustments may occur, giving 
rise to conditions — such as local heating and energetic deuterium motions — that are 
relatively more favourable for some of the D— D fusion mechanisms proposed in the 
literature. This description also suggests that the results of the electrochemical experiments 
could be very much sample — dependent. 

Let us next turn to the enthalpy release. Actually there are several adsorption, 
absorption and desorption processes involved in the experiment using an electrochemical 
cell. But here we will concentrate only on the heat of formation of PdD x , which has not 
received enough attention. We first note that in electrolysis using Pd as a cathode, because 
of its special ability to dissolve H/D, one generally does not observe any evolution of Hj/Dj 
at the beginning, as these are .absorbed by Pd. Also we may bear in mind that the values of 
the enthalpy release in the formation of PdD* quoted above from ref, 4 relate to the 
situation with molecular D 2 . If we were to measure these using nascent (i.e. atomic form) 
deuterium:, then the enthalpy released would be larger by the dissociation energy 8 of D 2 , viz 


C 1 

106 kcal per mole of D 2 . Assuming that the (cathodic) current in the Fleischmann-Pons 
experiment is only due to flow of D + , we can then infer the amount of deuterium impinging 
on the Pd cathode and then roughly estimate the rate of enthalpy release due to deuteride 
formation. We estimate that it is of similar magnitude as the claimed ’excess heat* by 
Fleischmann and Pons and emphasize the need to include it in the total energy balance 

To sum up, we feel that the neutronic signals reported to have be|(Aeln in some of 
the recent electrochemical experiments deserve to be viewed ift ih# the materials 

science of palladium deuteride. Although, if it is finally confirr&eH7"This so called ’cold 
fusion’ would be physically very interesting, the possibility that Jb;wiljlead to a significant 
new energy source appears doubtful at present. ri 

# %, J 

We acknowledge helpful discussions with a large number of colleagues especially Dr. 
P. K. Iyengar, Dr. S. Gangadharan, Dr.T. P. Radhaknshnan and Dr. S. K. Sikka. 



1. M Fleischmann and S Pons, J. Electroanal. Chem. 261, 301 (1989). 

2. S E Jones et al, Nature (London) 338,. 737 (1989). 

3. R L Garwin, Nature (London) 338 . 016 (1989). 

4. DM Nace and J G Aston, J. Am Cheirf6oc 79, 3619, 3623, 3627 (1957). 

5. G Nelin, Physica Stat Sol (b)45 . 527 (1971). 

6. R G Weast (ed), CRC Handbook of Physics and Chemistry, (CRC Press, Florida, 

1987), p.F 171 . fh * 

* Published in ’’Current Science”, June 5, 1989; Vol. 58, No. 11 pp 597—598. 


C 2 

B.A.Dasannachaxya and K Ft. R&o 
Nuclear Physics Division 

In this note we wish to bring to attention certain processes which have been invoked 
to explain solid state and chemical phenomena and which may have relevance to cold 

fusion 5 ' 4 . ® ~ 

Starting with Brownian motion in a liquid we note that the typical momentum 
gained by a foreign Brownian particle, of 1 size, is hundred times the average momentum of 
the molecules of the liquid 5 . This is because the momentum gained by the particle is not 
due to individual impacts from single molecules but because of collective motion of a 
number of molecules. We also note that the frequency of the random jumps tells us that 
such fluctuations in collective motions occur quite fregjiently over normal observational 
time— scales. Larger momentum transfers would also be imparted but lass frequently. 

In a solid, similar fluctuations may L*e less frequent, but larger 
may be expected especially in a system witfi j arge anharmoni city ; D motion in PdD* is 
known to be very anharmonic. The phase diagram of Pd— D system has been extensively 
studied and its similarity with gas— liquid transition has been discussed at length by 
Alefeld®. In the phase, with D/Pd ratio % 0.6, deuterium behaves like a lattice liquid with 
fairly fast diffusion (D -IQ' 7 cm^/sec. at -50°C in phase). Lattice Brownian motion has 
been invoked in the past to explain hydrogen diffusion in metals. The important difference 
in the present case is that the fluctuations of interest to us now are over atomic length 
scales (several angstroms) and times are typically lattice vibration times (picoseconds), as 
against hydrodynamic time scales of Brownian motion referred to earlier. If such large 
momentum transfers as those observed in Brownian motion (i.e. 100 — 1000 times mkT ) 
could be imparted to individual deuterium atoms over very short time scales, one can 
visualize a situation with a’ rauch improved chance of a nucleax reaction. 

There are a number of solid state phenomena 7 which give indication that this could 
indeed be happening. They have been discussed at length by Khait in a number of papers 
(see Khait 7 for Preview ). He has argued that there are several phenomena-in chemical 
and solid state physics— to describe which, equilibrium concepts are ‘fundamentally 
qualitatively' inadequate. The phenomena are mostly connected with rate processes 
involving a large activation barrier. For example, if one chooses to describe the diffusion of 
Ba in BaO in the temperature range of 1350 to 150Q°C by an Arrhenius type of equation, 
D = Do exp (- EA/kT), one obtains EA to be about 12 eV which is about 100 times kT. 
Similarly the activation energy for diffusion of Si, Ge or Bi in Si is -5 eV = 6xl0 4 Kelvin. 
In spite of such large activation energies measurable diffusion rates are observed. Such 
rates cannot be explained by assuming small thermodynamic fluctuations at the relevant 

temperatures. It was therefore proposed that short— lived large energy .fluctuations of a 

small number of particles is necessary to be invoked for explaining these phenomena. The 
mode of formation and dissipation of these fluctuations is discussed at length in reference 7. 
The basic point relevant to our discussion is that fluctuations involving large energies over 
short times do occur in solids. They can occur under ‘normal* conditions, but, the 
probability of getting such large energy fluctuations increases especially near certain 
transitions, under the presence of gradients (pressure, concentration, temperature, field) 
etc. It is also enhanced in the presence of high anisotropies and near the surface, in a 
direction perpendicular to the surface 7 . In all the ‘successful* cold fusion experiments 
reported to— date, electrolytic 1 ”' 2 , 4 and absorption/desorption 5 " 4 type, these ‘favourable* 

conditions exist. These fluctuations could impart energies which are 100 to 1000 times kT 


to the D atoms in PdD x . Thus, instead of the typical optical phonon energy of 50 meV, 
the deuteron can get energies ~50 eV due to the presence of a short-lived large energy 
fluctuation. It can then approach another deuterium to within -.22 A® assuming deuterons 
are considered to be in the presence of a uniform electron background (Screening length 1 
A Et = 1 eVJ. The frequency of reactions thus induced will depend on the frequency with 
which the fluctuations are produced in the solid. It is not possible at present to give a 
realistic number for this, calculated ab initio. _^p 

In conclusion, we note that a number of known phenomena give evidence for the 
presence of large energy fluctuations of small number of atoms oyen^ort times and these 
may be considered as one of the possible modes for ‘starting 1 a cold fusion which would 
then be followed by more nuclear reactions involving the nf?od&tefc of the first fusion 
reaction. Such fluctuations may vary considerably depending'pn the condition of the solid/ 
experiments, like concentration gradient, field gradient, nearness to a transition etc. 








ftp* ", 

M.Fleischmann and S.Pons, J .Electroanal.Chem: J61. 301-308 (1989). 

S.E Jones, Nature 338, 737-740 (1989). 

A de Ninno, Europhys. Lett. Preprint (1989), 

P.K. Iyengar, ICENES V (1989) and Private Communications of experiments at our 
Centre. % 

D.K.C. McDonald, Noise and Fluctuations (J. Wiley, NY 1952). 

G. Alefeld, Ber. Bunsen. Phys. Ghem. 76, 746-755 (1972). 

Yu.L.Khait, Physics Reports 99, 237-340 (1983). 

C 3 



S.N, Vaidya and Y.S. Mayya* 

Chemistry Division 

^Health Physics Division <*^3 

. C/$ 

a '4L # 

It is well known that the quantum-mechanical calculation of the fiSion rate for Dj 
molecule in free space yields a negligibly small value 1 of ~10" 70 s'* recent electrolytic 
experiments by Fleischmann and Pons 2 and by Jones et al 3 , how^vea 'suggest a possibility 
of a much higher fusion rate of deuterium in palladium at room temperatures. In this 
communication we propose that if cold fusion is indeed a rb||ffey{ it may be explained by 
the combined screening of the Coulomb interactions by itinerant deuterons and the 
conduction electrons in metals. The chief assumption of- our model namely, that deuterium 
exists as mobile ionized species in palladium under electrolytic conditions, is based on the 
following facts. It has been established from the electromigration measurements 4 that in 
palladium, hydrogen exists as a proton with an effective charge s+L Likewise, deuterium 
should also exist as deuteron with an effective tji&rge $3+1. Besides, deuterium is known to 
be highly mobile in palladium 5 . Under these’ ‘'^onchtions, the deuterons would participate in 
screening the electrostatic interactions ajong With the conduction electrons of the metal. 
The combined screening reduces the Coulomb barrier between the deuterons separated by a 
distance r by a factor exp(— kr) (where k k the screening constant), thereby leading to am 
enhancement in the fusion rate, " 

For nominal composition’ PdD with a substantial fraction of deuterium atoms in an 
ionized state, the condition df charge neutrality in the bulk can be written as 

Up + n<j — n c 

( 1 ) 

where np, n<j, n c are the bulk densities of the fixed palladium ions, mobile deuterons and 

the conduction electrons respectively. 


For calculating k, one sets up the Poisson's equation for the potential at a radial 
distance r from a test charge in terms of the net charge densities induced at r. The induced 
charge densities are obtained using Fermi—Dirac statistics for electrons and Bose— Einstein 
statistics for deuterons. Upon linearizing the densities as a function of the potential and 
using the Fourier transform techniques for the Poisson’s equation, it can be shown that 

k 2 = k c 2 + k d 2 (2) 

kc 2 =r 6 e 2 n c /E c (3) 

k<j 2 = [4 ire 2 nd/kbT] [zg’ 3/ 2 ( 2 ) /g3/ 2 ( 2 )] (4) 

In the above equations £<; = (2m c )' 1 (3r 2 h%c} a 43 is the Fermi., energy of the electron 
gas, kb is the Boltzmann constant and T is the temperature. The fugacity z (0<*<1) is 
related to the chemical potential via the expression z = exp(— p/kfaT) and. the function 
= [n” 3 /2 z n ] is well known in connection with Bose condensation 6 . Equation (4) is 
valid for temperatures T>T C where 

T c = [27rh 2 /mdkb] [nd/2.612' 


( 5 ) 


C 3 

The Bose condensation temperature of the ideal deuteron gas in PdD calculated from eq.(5) 
is 6.65K. Prom eq,(4) it follows that at high temperatures a T _l 72 . At low temperatures 
kj can diverge as (T— T^" 1 n as T~*T C \ Thus the contribution to screening from charged 
bosons is strongly dependent on temperature and exhibits a power— law divergence at T c + . 
However, this may not be realizable in PdD due to a sharp drop in the mobile fraction of 
deuterons at cryogenic temperatures. 

The Thomas— Fermi screening by conduction electrons is well known in solid state 
physics 7 . The calculations based on (2)— (4) show that the combined screening by electrons 
and deuterons is more effective than that due to electrons alone. 

To calculate the fusion rate in the presence of the combined' screening, we assume 
the existence of pairs of deuterons in the interstitial sites. of palladium lattice. These form 
metastable D 2 + ions or D 2 molecules which undergo spontaneous fusion at a rate 

R = A[#0)P, A = 2xl0-f pn 3 |-t (6) 

where ip(0) is the molecular wave function at the origin. At an internuclear separation r 

^(r) = [pw/8*- 5 n r*h] exp[-/?(r)] (7) 

«r) [2 1 2/i {E v - V(x)> - 1/4*»|>« - (l/x)]dx (8) 

where fi is the effective mass of the d— d system, w is the vibrational frequency, E v is the 
ground state energy, V(r) is interaction potential between the deuterons and r is the 
inner turning point on the potential energy curve (figure 1). To take the screening into 
account, we have used Jke potential 

-fi Jp' % 

= (l/ r ) exp (-kr) - 1.5/ [l+0.127r) 4 (9) 

which, for k = 0 is the same as the potential used by Jackson 8 for an unscreened D 2 + ion. 
All the quantifies' in the above equation are expressed in atomic units. In the present 
calculations ; vye"feave assumed that almost all the deuterium atoms are ionized and these 
can be extended to account the actual fraction of deuterons present by using appropriate 
values for nd“'and n c . 

The calculations using the above equations show that the fusion rate for D.j + ions 
which is 5 X 10~ 78 s' 3 in free space increases to 4 X 10~ 46 s‘ 3 in palladium when the 
screening due to the conduction electrons {k c - 1.0 au) alone is applied. On the other 
hand, under the combined screening by deuterons and electrons at 300K (k = 11.6 au), the 
fusion rate for D 2 + ion increases to 5 X 10” l6 s -1 . Similar calculations were performed for E >2 
molecule with the potential V(r) = (l/r) exp(-kr) — 2.18/ [1 + 0,18r] 4 , The fusion rates 
obtained were 1 X 10" 63 s' 1 for k = 0, 3 X 10~ 40 s’ 1 for conduction electron screening (k c 
~ l.Oau) alone and 1 X1CT 14 s' 3 for the combined screening by deuterons and electrons at 
300K (k = 11.6au). Thus it is seen that the spontaneous fusion rate increases to 10" 18 s” 4 for 
D 2 + j 011 to 10" 14 s” 1 for D 2 molecule at 300K, which lie in the range of values indicated 
by the cold fusion experiments. 

The fusion rate is a sensitive function of k<j which is proportional to np n . Under 
electrolytic conditions, the density of deuterons is higher on the palladium surface than in 
the bulk due to the existence of a large concentration gradient at the surface. Moreover, the 
mobility of deuterons in the surface layers will also be much higher. As a consequence, the 
fusion rate will be substantially higher in the surface layers than in its bulk. 


G 3 

Suppose an additional electric field is applied along the length of the 
deuterium— saturated palladium cathode during electrolysis in the Fleischmann and Pom 1 
type cell. Due to electromigration under the applied field, the fraction of the mobile 
deuterons will be increased. Further, as pointed out by Wipf 4 , a potential difference of say 
500 mV can create a deuteron density ratio of 10 s between the two ends of the cathode. 
Though this factor is probably an overestimate (due to space-charge effects)^jmder the 
additional electric field, n<$ at one end will nevertheless be enormously higher taaSJ that in 
the absence of the field. As a consequence, we expect that under the appliradton of the 
additional electric field, the screening will be substantially enhanced wjngh%wbdld, in turn, 
lead to a large increase in the fusion rate. Similarly an increase in the fusion rate can. be 
expected in the Ti— D gas experiment 9 by the application of the elec trie? field across the 
titanium metal. 

We thank Dr. V.K. Kelkar for computer calculations. We. axe indebted to Dr. J.P. 

Mittal and Dr. K.S.V. Nambi for their encouragement and support during this work. 



1 C De W Van Siclen and S E Jones, J.Phys. 12, 213 (1986), 

2 M Fleischmann and S Pons, J. Electroanal'Chem.- 261 . 301 (1989). 

3 S E Jones, E P Palmer, J B Czirr, D L Declek, G L Jensen, J M Throne, S F Taylor 
and J Rafelski, Nature (London) 338 . 737 (1089). 

4 H Wipf m Topics in Applied Physics: Hydrogen in metals I and II, ed. by 
G Alefeld and J Volkel, p 273, 300 (Berlin: Springer, 1978). 

5 J Volkel and G Alefeld in Topics in applied physics: Hydrogen in metals I and II, 
ed. by G Alefeld and J Volkel, .p 321 (Berlin: Springer, 1978). 

6 K Huang, Statistical Mechanics, p 262 (New York: John Wiley, 1963), 

7 C Kittel, Introduction to Solid State Physics, 5th ed. (New York: John Wiley, 1976). 

8 J D Jackson, Phys.B 4 v. 4 m 330 (1957). 

9 A De Ninno, A Frattqjillo, G Lollobattista, L Martinis, L Mori, S Podda and 
F Scaramuzzi, Frgpfjnt (1989). 

•‘Modified Version of the paper Published in 
Pramana— J.Phys., Vol.33, No. 2, pp L343 — 346; 1989. 


C 4 



Swapan. K. Ghosh, H K. Sadhukhan and Ashish K Dhara 4, 

Heavy Water Division 
^Theoretical Physics Division 

An efficient means of achieving fusion of isotopic hydrogen nuclei has only remained 
a long cherished dream in the field of nuclear science due to the need for high energy 
acceleration to overcome the associated Coulomb barrier. The recent discovery of com 
nuclear fusion by Fieischmarm et al 1 and Jones et al 2 during electrolytic deposition and 
consequent compression of deuterium gas into palladium or titanium electrodes, has thus 
generated great current interest 3 . The fundamental theoretical problem posed by this 
spectacular phenomenon is to explain the reported fusion rate (~10" 2 ° per deuteron pair per 
sec), which is many orders of magnitude larger than that calculated 4 (~1G~ 70 per deuteron 
pair per sec) for subbarrier fusion involving conventionaLquantum mechanical tunnelling of 
the Coulomb barrier in a deuterium molecule. 

Cold nuclear fusion with even mucJh*Meheh fusion rate is however well known in 
muon-catalyzed fusion (see, for example, Jones*), where the enhancement of the tunnelling 
probability arises from a drastic reduction of'the internuclear distance due to replacement 
of an electron in a deuterium molecule by a massive particle like muon 6 . An attempt has 
been made to extend this concept for explanation of cold fusion in a metal lattice by 
considering the binding particle to be a quasi electron of an effective mass m* higher than 
the electron mass m. The reported tunnelling probability' can be predicted 7 using a value of 
(m*/m)» 5, which is however too high to be realized, 

In the present work, we propose a new mechanism for the observed cold fusion in 
deuterium-loaded palladium.' We believe that a consideration of mere tunnelling in an 
isolated deuterium molecule is not adequate to explain the observed cold fusion. The 
central theme of our approach is that the fusion in a metal lattice is guided by a collective 
phenomenon resulting into a screened Coulomb barrier which leads to this drastic 
enhancement o( t the barrier penetration probability. 

Extensive studies have been reported on various aspects of hydrogen in metals 8 and 
the palladium— hydrogen system m particular 8 . It is known that the deuterium gas 
compressed inside a palladium lattice dissociates into atoms 13 * 12 , each carrying a net 
positive charge close to unity fife). Also the quantum mechanical zero point motion of 
these D + ions (which axe bosons) has large amplitude and thus there emerges the picture of 
a mobile delocalized charged boson gas of D + ions, with the effective binding potential 

provided by the system, of electrons and the lattice. For simplicity, m this work we consider 
a jellium type model in which the positively charged boson gas of uniform D 4 ion density is 
subjected to an .attractive potential from the lattice and the electrons, represented by a 
smeared out uniform negative charge distribution. The resulting system is an example of a 
quantum plasma of bosons and would be .associated with plasma oscillations leading to 
screening of the interaction potential., (For plasmon screened, enhancement of 
thermonuclear fusion rate in stellar interiors, see Refs 13 and. 14). The D 4 ions now interact 
with an effective screened Coulomb potential of the form 

V e ff (r) = (e 2 /r)exp(— r/1), (1) 

characterized by the screening length /. The screening parameter / is of the order 15 , 16 of /« 
(ve/wp), where v 0 is the single particle velocity of D + ions and w p is the plasma frequency 


C 4 

given by 

w p * (4 t n<j e 2 /iRd) I/3 

( 2 ) 

with n d , e and md denoting the number density, charge and mass of the D* ions. For a 

classical plasma, vq is obtained from the average thermal energy, while foj^fermions, it 
corresponds to the velocity at the top of the Fermi distribution. For the present boson 
system, we assume the occupancy of only the ground state, i.e. consider a fp?s' temperature 
and calculate vo by equating (1/2) ra^vo* to the ground state energy ^ 

*o = (h 2 /2md) (3.142) 2 / r s 2 

i/' ^ 

of a D + ion in a sphere of radius r s , defined in terms of the msah-.Volurne of the ion in the 
metal, by 

(4/3) »r r s 3 = n d ‘ 


‘ -Vi 

i § 

The screening length l can thus be expressed as 

l a (3. 142/3* n ){mfc(&pv (rjao) 1 n ao, (5) 

where ao is the Bohr radius for the electron. The effective screened Coulomb potential in 
which the IT ions move is thus defined'by eqs (1) and (5). The corresponding Schrodinger 
equation for a pair of D + ions can be written as 

f(d 2 / dr 2 ) + { 2(/j d /h 2 ){ B-V «ff(r)} - J(J+l)/r 2 } *(r) - 0 (6) 

where the centre of map motion and the angular dependences have been separated out. 
Here,x (r) = r# (r) denote the radial wave function, p r j (=m d /2) is the reduced mass and 
E is the energy of relative motion of approach. In order to employ the semi-classical WKB 
approximation to thist raxlfaJ equation, the term 3(J+1) in eq. (6) is replaced 4 by (J+l/2) 2 . 
For J = 0 state, this involves the addition of a centrifugal barrier term (h 2 /m d )ir 2 to the 
effective potent^ (r) . The fusion rate is determined by the ratio of the probability 
densities x 2 (rit);€ud X 2 ( r a)> where r n is the distance of closest approach for fusion (equal to 
the nuclear radius) and r a is the classical turning point determined by the zero of 
’ as % 

Q 2 (r) = (2p d /h 2 ) {V eff(r) - E} + l/4r 2 (?) 

The transmission coefficient T(E) for this barrier penetration is given by the well known 
WKB approximation as 

T(E) = exp [-2 /!* | Q(r) |dr ], (8) 


which for convenience can be rewritten as 

T(E) = (r n /r.) exp[-/f* (2 IQWI - r-')dr ] , (9) 


The fusion rate (per deuteron pair per sec) is now obtained from the relation 

P = Sv 0 T(£) (10) 

where vq is the frequency of attack representing how often the two D* ions approach each 
other and can be calculated by writing hvQ » B, the energy of relative motion (- ep)* The 


C 4 

parameter S is a selection factor for a particular channel of D + D nuclear reaction 

We have carried out the integration of eq. (9) numerically and calculated the fusion 
rates at several values of D:Pd ratio. The results obtained by using the density of Pd metal 
as 12.16 g/ml and a value of 10" 13 cm for the huclear radius r^ are shown in table L For a 
normal D:Pd ratio of 0.6, one has the number density n<j = 4.13 x 10 22 /ml; thg screening 
length 1 = 0.055 ao and the frequency factor i>$ = 7.68 x 10 n /s. For an S factor of unity 
(which is used for all the present results), the fusion rate is 3.68 x 10" 20 p^p%uteron pair 
per sec and it increases to 1.07 x 10‘ s8 for a D:Pd ratio equal to unity that 'corresponds to 
high pressure. These preliminary order of magnitude calculations thus lead to results which 
compare very well with the reported experimental cold fusion rates. Calculations using 
more exact expressions for the screening length or correct value for the S factor (less than 
unity) although would modify the numerical results somewhat fi qualitative predictions are 
not expected to change. iNfcS 

The present study reveals that the formation of charged boson plasma state of 
deuterium in the palladium lattice and the consequent new mode of screening of the 
interaction of deuterium ion pairs might be responsible%r the tremendous enhancement of 
the fusion probability via quantum mechanical tunnelling. Our work differs from most of 
the recently reported calculations where exi^t^n^exaf deuterium molecule is assumed and 
either the concept of effective electron mass 17 of electron (fermion) screening 18 has been 
invoked. It also indicates the role of palladiumdn cold fusion since in this metal deuterium 
is present as positive ions, whereas in other metals the charge on each deuterium atom 
might be zero or even negative in some Cases 10,12 . 

In conclusion, we emphasize that the screening mechanism due to a quantum boson 
plasma can significantly enhance the cold fusion rate. However, due to extreme sensitivity 
of the predicted rate on the screening length, further studies incorporating the effect of 
temperature, non adiabatic degrees of freedom, non equilibrium conditions etc are in 
progress to confirm the rBie qf the suggested mechanism in the observed cold fusion. 


1 M Fleischmann, S Pons and M Hawkins, J. Electroanal. Chem. 261. 301 (1989). 

2 S E Jones et al, Nature 338 , 737 (1989). 

3 J S OoheSband J D Davies, Nature (London) 338, 705 (1989). 

4 CD Van Siclen and S E Jones, J.Phys. G12, 213 (1986). 

5 S E Jones, Nature (London) 321 , 127 (1986). 

6 J D Jackson, Phys.Rev. 106, 330 (1957). 

7 R L Garwin, Nature (London) 338 616 (1989). 

8 G Alefeld and J Volkl (Eds), Top.Gurr.Phys. Vol 28 and 29 (Berlin: Springer, 1978). 

9 FA Lewis, The Palladium Hydrogen System, Ch. 10 (New York: Academic, 1967). 

10 J Friedel, Ber. Bunsenges. Phys.Chem. 76, 828 (1972). 

UFA Lewis, Platinum Met. Rev. 26, 20, 70,121 (1982). 

12 E G Ponyatovsky, V E Antonov and I T Belash, Problems in Solid State Physics, 
Edited by A M Prokhorov and A S Prokhorov (Moscow: Mir, 1984). 

13 EE Salpeter and H M Van Horn, Astrophys. J. 155, 183 (1969). 

14 H L Duorah and A E M Khairozzaman, Pramana — J. Phys. 15, 145 (1980). 

15 D Pines, Elementary excitations in Solids, Ch. 3 (New York: Benzamin, 1963). 

16 S Raimes, The Wave Mechanics of Electrons in Metals, (Amsterdam: North 
Holland, 1963) p.289. 

17 S E Koonin and M Nauenberg, Nature (London) 339, 690 (1989). 

18 C J Horowitz, Phya Rav O (1989), in Preaa. 


C 5 


T G K amhi k, M Snmvasan and A Shyam 

Neutron Physics Division „ 

The reports on detection of neutrons and tritium above bacJqor^wHIevels in 
deuterium gas-ioadftd titanium and palladium targets 1-2 as well as ijj, .|f : e||l&^tic cells 3-4 
using these metals as cathodes immersed in electrolyte containing he&wfwraSt (D 2 O), have 
led to the concept of the so called cold fusion i.e the occurreiic^|pPi^preciable fusion 
reactions among deuterium nuclei at room-temperature. Several theoretical speculations 
have been put forward, since the beginning of the 'cold fusion epjE' tep'ex plain the existence 
of such a process. However almost simultaneously it has been^ Suggested, as for example at 
the AlP meeting 5 held at Santa Fe in May '89, that the fusion may not be 'cold 1 as such 
but might be resulting from some kind of beam— target interaction of accelerated particles 
or hot micro plasma formed at certain sites in the^ciids due to fracture etc. This 
proposition, termed as 'fracto— fusion', is based on tHejgPast observations of interaction of 
gases with metals and emissions of all kinds of, particles and radiations from cracks or 
cleaved surfaces of crystalline solids. This u&te presents a brief survey of this type of 
studies with a view to assess the role of sucfrjpechanisms in 'cold fusion' and plan possible 
experiments to confirm them. >. 

in solid state science, the process, of emissions emanating out of mechanically 
deformed crystals, specially thin oxidfe, layers, has been known for a long time 6 m the form 
of ‘tnbcluminescence 1 where light is emitted on application of stress to the crystals. Several 
kinds of emissions including^ ejectyofcs, negative and positively charged ions, are found to 
accompany such process 7 ,,., ■ bubsuquent studies by Dickinson and others workers® have 
correlated these emission! ..i® Lite appearance of cracks in the materials m shown by the 
recorded acoustic signals. '1$jisides charged particles even neutral gas is found to evolve in 
some cases 9 . A burst, of as many as 1CH particles 7 may occur in a short duration of 10/us 
which is the time a crdek of the size of a few tens of /im would take to propagate (with the 
speed of sound) although it may continue for a few minutes at lower rates of emission 10 . 
Origin of this phenomena has been attributed either to the thermionic emission 7 generated 
due to very short lived microplasma formed at the crack sites or to the field emission 7 
resulting from 'strong electric fields, developed due to charge layer separation, Some 
independent estimates of temperature at these sites indicate an appreciably large value 11 of 
about 1 eV giving credence to a thermionic mechanism. At the same time it has also been 
fairly well established that the sites of cracks or cleaved surfaces of tome crystals can result 
in transient electric fields of more than 15 KV /cm 12 ' 13 . Sources of such transient fields are 
attributed to micro— and macro-cracks, besides the clusters of dislocations 53 . All these 
experiments have been mainly earned out with insulating crystals and as pointed out by 
Komfeld 12 , due to neutralization by surrounding gas, it may not be possible to sustain a 
greater (than 15 KV/ cm) field although a much higher value of 10 6 — 1G 7 V/cm has been 
quoted by Klyuev et al 14 Another source of this Tracto—e mission 1 could be a shock— wave 
or a stress— wave created by the crack tip motion which may give rise to electrical fields as 
suggested in ref 10. In brief, as pointed out by Dickenson 1 ®, fracture of insulators can result 
in production of highly localized heat, creation of excitations and defects, emission of 
excited and .reactive species in gas phase, separation of charges on crack wails accompanied 
by intense electric fields and production of acoustic waves. 

In the context of 'cold fusion 1 phenomena, above facts need to be viewed along with 
the well known phenomena of large cracks found to develop in metal— hydrides and their 
embrittlement on hydrogenetaion 15 An observation by Dickenson 1 ® oft the multiplicity of 


C 5 

the number of electrons emitted in a burst from crack sites in crystals could be very 
significant to the suggested hot fusion mechanism. One of the evidences in ref .2. , analysed 
in details elsewhere 5 ^ about the non— poisson distribution of neutrons generated from the 
cold fusion experiments is quite similar to that of the above mentioned electron— emission 
which is also found to occur in singles as well bunches of many at a time. The mechanism, 
of ’fracfco— fusion 1 2 has been widely quoted at the Santa Fe meeting of AIP oifaeold fusion 5 as 
also by Menlove 58 and Levi 1S . . - 

The experiment, of Klyuev et al 14 in 1986 where a striker -waas^cc^ferat ed by a gas 
gun to about 200 m/s and impacted onto a LiD crystal is a ckwab^Jms direction. With 
proportional counters, whose output was analyzed through a pulse height analyzer, neutron 
signal significantly higher than the background was obtained/ wpi an estimate of about 
10 neutrons per fracture of the LiD crystal. Since this ex^ertfeeM was basically conducted 
to check the possibility of the above mechanism i.e. fusion caused by deuterium ions, 
accelerated through large electric— fields generated at Ijje crystal’s cleaved faces, the 
positive signal of neutrons appears to corrobora tfeLri . Sietz 20 has also proposed the 
possibility of hot fusion in which he indicates ^^•-‘iitgh-temperature (about 20 eV) 
microplasma might be created at certain sites utau^Erystal due to the formation energy of 
D 2 molecule alone else through accelerated ions fr conditions favour. 

'*%. 'V 

Jr VV 

In a recent experiment at BrooSfr%yeB Lab 23 , bunches of up to few hundred D^O 
ions, accelerated to about 325 Kev enprgies-wiih an electrostatic accelerator and impinged 
on deuterated titanium target, are reported to have resulted in yields of 0.1 neutron/ 
second — na of ton current. With various control experiments, the origin of these reactions is 
suggested as due to cluster— ion impact on TiD target only. The reaction— rate though quite 
small still shows several orders , of magnitude higher value than might be expected on the 
basis of usual calculations,.. Thus discrepancy is sought to be explained by the authors on 
assumption of certaiy'cofapiession and heating due to impact of cluster —ions in the 
reaction area. It is p<%jfcl^that this type of conclusions, if studied further, would also help 
in better analysis ^aaw^mderstanding of the probable mechanisms responsible for the 
observed cold f 1 isi axi p h e n o m e n a . 

Ikayg amdlMiyamaru 22 too have tried to investigate if stress in materials could give 
rise to the observed fusion. In their experiment, an electrolytically charged s.s. electrode, 
sputtered with palladium was mechanically bent to generate stress in it but the detectors 
did not show any tritium or neutron signal, although significant amount of excess chemical 

heat was found to be generated. 

Whether the ’fracto— fusion 1 could really be the process responsible for the fusion in 
deuterated metals can only be established by further experiments. For this purpose, it may 
be possible to accelerate a deuterated projectile to a few km/e with the help of an electric 
gun or railgun and then impact it against another similar target. The main problem, in 
detecting the expected low neutron counts with these accelerators is the accompanying 
electromagnetic noise. Meanwhile along with fusion signals, the detection of electric fields 
aod related acoustic emission during cracking in the present type of cells or gas-loaded 
metals, should be able to indicate if the fracturing of metals because of deuterium loading 
ss at least partially, if not completely, involved in the observed cold fusion phenomena. 

1 A De Ninno et al, "Evidence of Emission of Neutron From a Titanium— Deuterated 
System”, Europhys. Lett* 9, 221 (1989). 

2 P K Iyengar, "BARC Cold Fusion Experiments”, Paper presented at the Fifth Inf, 
Conf. on Emerging Nucl. Energy. Systems, July 1989, Karlsruhe (FRG). 


C 5 

3 S E Jones, "Observation of Cold Fusion in Condensed Matter", Nature (London) 
m, 737 (1989). 

4 M Flekchmsn and S Pons, " Eiectrochemicaliy Induced .Nuclear Fusion of Deuterium”* 
J Etectroanal. Chetn. Ml, 301 (1989). 

5 Robert Pool in Science 244, 1039 (1989). 

6 A j Walton, "Triboiumemscence", Adv. Phys. 26, 887—948 {1977). 

7 B Z IkwenbkiiB et al, "Spontaneous Emission of Charged Particles h Photons During 

Tensile Deformation of Oxide— Covered Metals under Ultra * raijh~V&cuiiitt* , 
j. Appl Phys. 4§, 5262-73 (1977). , 

8 J T Dickinson et al, "Acoustic Emission & Electron Emission During Deformation of 
Anodized Aluminum", j Vac. So. Tech. V’ 429—432 (1980). 

9 1 A Larson et ai, "Emission of Neutral Particles From Anodized Aluminum Surfaces 
During Tensile Deformation", j. Vac. So. Tech. i§, 590-591 \ 1979). 

10 J T Dickinson et al, "The Emission of Electrons h Positive Ions From Fracture of 
Materials", J. Mat. Science 16, 2897-2908 (1981) 

11 P G Fox et al, "Thermal— Induced Thermal Decomposition in Brittle Crystalline 
Solids”, Proc Roy. Soc. Lond. A317 . 79-90 (1970k 

12 M I Kornfeld, "Electrification of Crystals by Cleavage”, J. Phys. D. 
Appl. Phys... 11, 1295—1302 (1978), 

13 Y I Golovin et al, "Transient Electric.* Pkld of a Fast Cleavage Crack in LiF Single 
Crystals", Sov. Phys. Sol. State 27, 671—674 (1985). 

14 V A Klyuev et al, "High-Energy Proelkes Accompanying the Fracture of Solids", 
Sov. Tech. Phys. Lett. 12, 551-552 (1986) 

15 eg, J P Bleckledge in Metal Hydrides ed by W M Mueller, J P Bleckiedge and 
G G Libowitz (Academic Pres$, NY, 1968). 

16 J T Dickinson et al, "Emission of Electrons And Positive Ions Upon Fracture of Oxide 
Films", J. Vac, Sci. Tech. 18 ;> 238-242 (1980). 

17 A Shyam et al, "Multiplicity Distribution of Neutron Emission in Cold Fusion 
Experiments", ThisfRiport, Paper A4 (1989). 

18 H Menlove as quoted by C Joyce in New Scientist No. 1671, p 34 (1989). 

19 Barbara G Levi in Phys. Today, June 1989, p 19. 

20 R Sietz, "Fusion in From Cold?", Nature (London) 33S, 185 (1989). 

21 R J Beuhier "Cluster —Impact Fusion", Phys. Rev. Lett. 63, 1292—95 (1989). 

22 M Ikaye and II Miyamaru, "Chemical Heat Production of Pd Electrode Electro- 
Chemically -Charged With Deuterium & Hydrogen", Chem. Exp. 4. 563—566 (1989). 



Most of the papers included in this report were 
transferred into a personal computer directly from the typed 

-?v "*wa 

manuscript using a digital scanner followed by Optical 
Character Recognition (OCR) software.. This way we could avoid 

„ lyr M 

having to retype the papers all over again. The files were then 
passed through a spell ingi Cliecic program and then transferred 
into ” T 3 Scientific Word processor” before being printed out in 

• ■■ “ Xg, 

Proportional Roman Font (or ”Pica font” as it is called) « 

|/| * 

The Editors would like to place on record the significant 
help rendered by Sat. Padma Satyamurthy, Dr. A. Shyam, Shri. 
T.C. Kaushik and Kum. V. Chitra of the Neutron Physics Division 
in this onerous task. The Editors also thank Srat. Santha Nair 
and Shri, D.V, Periera of the same Division for their help in 
the word processing jobs. 

The Editors also wish to thank Dr. G. Yenkateswaran of the 
Water 'Chemistry Division for his help in preparing Table I of 

the StttiBary..